HEALTH CARE UTILIZATION AND EXPENDITURES

The U.S. health care system is unique among industrialized nations because it lacks a national health insurance program. The United States relies instead on private health insurance that individuals or companies must purchase. The public insurance system is limited to those who are aged or disabled (Medicare) and to some individuals who are poor (Medicaid). The delivery system for health care in the United States is almost entirely private, with only a small sector of government providers who primarily target the poor and uninsured population. Moreover, the U.S. health care system is largely unplanned and has limited regulation, even though the government is a large payer of services. In contrast, other industrialized nations have national health systems that provide coverage for their populations that are generally independent of employment (Blendon et al. 1995). These systems are more comprehensive and less expensive than the U.S. system, which excludes 43 million people (U.S. Bureau of the Census 1998b). This article examines the factors that have contributed to a system that severely limits access to millions of individuals in the United States.

We first examine health care utilization and expenditure patterns at the macro level. This approach is used by public policy makers at the federal, state, and local levels and by private insurers in decision making about the allocation and control of public and private resources for health care services. The second approach is a micro-level focus on the patterns of use and expenditures by individuals or groups of people in society and the factors that limit (or improve) access to services.

GROWTH IN EXPENDITURES

The most significant feature of health care in the United States during the l980s and l990s was its overall high growth rates and its total expenditures. Total national health expenditures increased by 283 percent between l980 and l990 (Smith et al. 1998). Although the annual growth rate in spending slowed from 11 percent in l990 to 4.8 percent in l997 (Levit et al. l998), expenditures were over $1 trillion in l996 and are projected to exceed $2 trillion by the year 2007 (Smith et al. 1998). Health care expenditures in 1997 amounted to $4,000 per capita (Table 1).

Health expenditures increased from 12.2 percent of the gross domestic product (GDP) to 13.5 percent in l997 (Levit et al. 1998), and they are expected to reach 16.6 percent of GDP in the year 2007 (Smith et al. 1998). The devotion of a large percentage of the total GDP to health costs is a concern because such dollars are then not available for other needs.

Factors Increasing Expenditure Growth. Levit and colleagues (1998) have examined three basic components in the growth of health spending. The first component is the economy-wide increase in prices and inflation (about half of the total increases in the l995–1997 period). The second is the change in volume and intensity of services. Levit and colleagues (1998) found that the volume and intensity of service use were about the same in the l990s (about one-third of the total expenditure increases in l997). Finally, there is medical price inflation over and above these other factors. In the l990–l995 period, excess medical prices were important factors—almost one-third of the total increases. Between l995 and 1997, this factor was only about one-sixth of the total increase (Levit et al. 1998). Even so, excess medical prices, or the "greed factors," are an important focus for policy makers because they represent the area where savings may be targeted.

Although health economists rarely focus on administrative costs, some scholars are now paying attention to the role of high administrative costs in the United States. For example, Hellander and colleagues (1994) estimated that administrative overhead accounted for 25 percent of total U.S. health system expenditures in l993. This included the costs of insurance overhead, hospital administration, nursing home administration, and physicians' billing and overhead. There is some evidence that such costs are increasing. The excess bureaucracy is related to the complex payment and billing systems that have been created in the United States. Consolidating the 1,500 insurers in the United States could reduce overhead costs.

Kenkel (1993) found that health maintenance organizations (HMOs) reported that 18 percent of their revenues were used for overhead and profits, compared with traditional insurance plans (14 percent) and Medicare (only 3.5 percent).

Economic analysts argue that the slower spending during the latter part of the l990s reflected several factors. First, where employers sponsored health plans, there was a shift of many workers from indemnity health plans into managed care (Levit et al. 1998), where premiums may have been lower. Second, managed care was successful in negotiating discounts with providers because there was an excess capacity of providers. Growth in managed care enrollment was found to have only a small effect on the volume and intensity of services. Finally, the general and medical-specific inflation rates declined (Levit et al. 1998).

Ginsburg and Gabel (1998) argue that spending was down in the l990s because of a sharp decline in provider revenues and payrolls. They show that hospital and physician spending increases in l997 were less than the growth rate in the GDP (which was 2.8 percent in l997). Although there was a slight increase in payroll costs between l996 and l997 in hospitals, most of the growth was due to increases in hours worked, and not in number of employees. At the same time, drug costs had increased by 11.5 percent, primarily because of increases in the use of drugs rather than the price (Ginsburg and Gabel 1998). Premiums for health insurance increased by only 3.3 percent between l997 and l998 (Ginsburg and Gabel 1998). The premium increases showed substantial profit margins in l992–l994, but lagged behind expenditures while profit margins in health plans dropped in l996–1997 (Ginsburg and Gabel 1998).

Health care spending increases are expected to continue because of the growing demand for medical services, continued economy-wide inflation, and medical price inflation. See Table 1. Most of the expenditure growth is expected to take place in the private sector, with lower spending in the public sector (Smith et al. 1998).

The high growth rate for health expenditures has had several consequences. First, the growth rate has dominated the attention of policy makers at all levels as they have attempted to struggle with cost containment. Most policy efforts have been directed toward reducing program utilization and expenditures. This focus on cost containment has set the national political agenda, relegating issues of access to health care and the quality of care to a second level of consideration. This agenda is of great concern because the number of uninsured is increasing and the quality of health care urgently needs to be improved (Chassin et al. 1998). Moreover, the high expenditure rates continue to be used as a rationale by some policy makers for why the expansion of public programs and a national health insurance policy are not financially feasible in the United States.

Growth Through Consolidations, Mergers, and Acquisitions. Overall health expenditures are increasing in part because of the changing structure of health care financing and delivery systems. Kassirer (1996) documented the rapid trends in consolidation of the health care industry, creating larger corporations during the l990s. Srinivasan and colleagues (1998) reported that, in 1997 alone, there were 483 mergers and acquisitions involving health services companies totaling $27 billion and 33 mergers and acquisitions involving HMOs totaling $13 billion. The big winners in mergers and acquisitions are the corporate executives of these large organizations, the stockholders, the lawyers who broker the consolidations, and the health care consultants. The losers are the physicians and nurses who may lose income and even jobs. It is not clear that these mergers and acquisitions are resulting in any overall savings to the system or in improvements in the quality of care. Rather, they appear to be increasing profits and market power (Kassirer 1996).

Srinivasan and colleagues (1998) reported that the growth rate in for-profit health service companies and HMOs has increased rapidly in recent years. Between 1981 and 1997, for-profit HMOs grew from 12 to 62 percent of total HMO enrollees and from 18 to 75 percent of all health plans. Between l987 and l997, there were 210 offerings for health services companies valued at $6.7 billion and 23 HMO public offerings for $1.4 billion. Health services capitalization was $113 billion and HMO capitalization was $39 billion in l997, showing dramatic growth rates. Stock prices for health care industry firms generally outperformed the overall stock market over the decade. As these changes occurred, there were also many conversions of nonprofit health care organizations to forprofit status in order to obtain capital and survive in a highly competitive environment. In l997, there were 81 conversion foundations with assets of $9 billion as a result of the conversions to for-profit status (Srinivasan et al. 1998).

Every year Forbes magazine lists the largest health care corporations in the United States and details their stock performance. Gallagher (1999) reported $10 billion in mergers and acquisitions in l998, including many in the HMO and long term care sectors. The managed care stocks were reportedly healthy in l998, climbing 15 percent over l997 (Gallagher 1999). HMO mergers and acquisitions were slower in 1998 than in the previous year. By comparison, hospital stocks were off 22 percent on average and long-term care providers were off 56 percent (Gallagher 1999).

What scholars and policy makers have generally thus far ignored are both the economic and the social costs of these changes in the structure of the health care system. The costs associated with mergers, consolidations, and profits are not identified in the analyses of health care costs (Levit et al. 1997). As long as the health industry is profitable, we can expect growth in the private profitmaking sector.

PRIVATE PAYERS

The payers of health services dominate the policy decision-making process in the United States. At the same time, the payers of health care have shifted over the past decade.

Private health insurance corporations dominate the health system because they have the lion's share of the resources and power in the health industry. For instance, private health insurance firms earned $348 billion in premium income in l997 in the United States (Levit et al. 1998). See Table 2. Private health insurance companies include both the traditional indemnity insurance plans as well as managed care plans. The private health insurance industry—including managed care organizations—spent $34.5 billion (or 10 percent) of their premium income in l997 on administration and profits.

Light (1992) has documented many of the problems with private indemnity health insurance plans. These plans have practice medical underwriting that excludes individuals with medical problems (e.g., chronic diseases) or raises the premiums for such policies to levels where individuals

cannot afford to purchase them. There are indirect approaches to reducing risk, such as waiting periods, copayments, and payment ceilings, along with the exclusion of certain procedures, tests, or drugs. Light (1992) also documented the problems with inaccurate or manipulative premiums and discrimination, which includes policy churning (switching policies before the waiting period ends) within group underwriting for those individuals at high risk, renewal underwriting (increasing the costs for those with new medical conditions), and selective marketing. There are many techniques to deny or delay payment for services. All these practices have negative consequences for individuals and groups seeking insurance. These problems are the result of a health care system geared toward profit and cost containment and of ineffective regulation of health insurance companies at the state level, although states do vary in their regulatory activities.

In l996, Congress passed the Health Insurance Portability and Accountability Act (HIPAA). This legislation, sponsored by Senators Edward Kennedy and Nancy Kassebaum, was designed to address some of the problems with health insurance by making it easier for persons who have lost employer-provided insurance to qualify for other coverage. However, Kuttner (1999) concludes that HIPAA is of little help because there are no regulatory controls on insurance premiums and no subsidies for those persons who need the insurance because most have lost their jobs and the ability to pay for the insurance. Similarly, the l985 Consolidated Omnibus Budget Reconciliation Act (COBRA) allowed people leaving employment to pay insurance premiums out of pocket for up to eighteen months to retain their coverage, but again, this is of little help for those without the resources to pay for the insurance (Kuttner 1999).

Managed care organizations (MCOs) had the most dramatic growth in health care in the l990s. As one study reported, "Managed care isn't coming; it has arrived" (Jensen et al. 1997). In l995, 73 percent of insured U.S. workers received their coverage through an HMO, a preferred provider organization (PPO), or a point-of-service plan (POS) (Jensen et al. 1997, p. 126). See Table 3. HMOs are health plans that provide care from an established panel of providers to an enrolled population on a prepaid basis. PPOs are health plan arrangements where networks of providers agree to accept plan payment rates and utilization controls. POS plans are HMOs where members may self-refer outside the established network but are required to pay deductibles and other costs. This means that conventional, or fee-for-service (FFS), plans have experienced a precipitous decline in recent years. Between 1993 and 1995, the percentage of workers in conventional health plans dropped from 49 to 27 percent (22 percent), while significant growth occurred in HMOs (6 percent), PPOs (5 percent), and POSs (11 percent) (Jensen et al. 1997, p. 126). Eighty-five percent of the U.S. work force was covered by some type of managed care plan in l997 (Levit et al. 1997).

There have been numerous studies of HMOs and their performance (Miller 1998; Miller and Luft, 1994, 1997). Davis and colleagues (1995) surveyed individuals in FFS plans in comparison to those in managed care. They found that many individuals (54 percent) had changed plans within the past three years. Of those who changed plans, 73 percent had done so involuntarily, either because of employment coverage changes or changes in jobs or moves. Of those persons in managed care who had changed plans, 41 percent also reporting having to change physicians involuntarily, compared with only 12 percent of enrollees in fee-for-service coverage.

Although most respondents were reasonably satisfied with their health insurance, those in managed care were more likely to rate their plan as "fair" or "poor" and were less like to rate their plan as "excellent." Those individuals in managed care plans reported less access to specialty care, less availability of emergency care, more waiting times for appointments, and less overall plan satisfaction (Davis et al. 1995).

Other research on the quality of managed care services points to concerns for populations with particular health needs. Miller (1998) found mixed results for access to care in HMOs among enrollees with chronic conditions and diseases. He found enough negative results to raise serious concerns about access. Nelson and colleagues (1997) found less satisfaction with the amount of services received among enrollees who used HMO services. For populations with a variety of needs, the concerns are also mounting. For example, studies of elderly ill persons revealed that HMO enrollees report poor quality of care and lower utilization than FFS enrollees (Clement et al. 1994; Shaughnessy et al. 1994). The problems that the elderly confront reveal spaces where government programs (i.e., Medicare) and private institutional practices (HMOs) need expanded coverage and improved coordination.

Between l989 and l995, the enrollment of Medicare beneficiaries in HMOs grew rapidly, so that HMOs covered about 10 percent of the total beneficiaries and the remaining persons received traditional FFS care (Welch, 1996). The Balanced Budget Act of l997 made a number of changes in the Medicare program, called Medicare+Choice, that were intended to expand Medicare beneficiaries' options (Christensen l998). These included: (1) PPOs, where networks of providers agree to accept plan payment rates and utilization controls; (2) provider-sponsored organizations (PSOs), which are plans organized by affiliated health care providers; and (3) medical saving accounts (MSAs) established by enrollees to receive the Medicare amount that would be paid to an HMO. Enrollees must pay all costs out of the MSA that would be covered by Medicare as well as out-of-pocket expenses (Christensen 1998). Congress expected that the Balanced Budget Act would accelerate the growth in HMO enrollment to nearly 34 percent by 2005. In spite of these changes to implement this new program, growth may be slower than projected because some HMOs are unwilling to accept Medicare members because they consider Medicare rates to be too low. In l998, HMOs announced that they were dropping approximately 400,000 of their 6 million enrollees because they consider the rates too low (Pear 1998).

There are two advantages of HMO enrollment for Medicare beneficiaries. The first is that premium costs have generally been minimal and the second is that some HMOs include prescription drug coverage and other benefits not covered by Medicare. Recently, however, drug benefits for Medicare beneficiaries have been reduced or capped by a number of HMOs (Kuttner 1999).

HMOs have also dramatically captured the Medicaid enrollment in many states. President Bill Clinton streamlined the process that allowed states to have waivers of the federal regulations that provide mandatory enrollment of Medicaid beneficiaries in HMOs (Iglehart 1995). The expectation was that HMOs would control Medicaid costs and encourage more coordinated forms of care by increasing the likelihood of having a primary care physician. The Kaiser Commission on the Future of Medicaid (1995) summarized 139 articles, books, and reports on Medicaid managed care and found mixed results. They found that the use of specialty services and emergency room care declined, but there was not a consistent increase or decrease in use of physicians' services or hospital care. The issue of cost savings was mixed, and quality and satisfaction rates were considered to be comparable. On the other hand, Iglehart (1995) documented serious problems with the implementation of Medicaid managed care in some states, including Florida and Tennessee.

Pauly (1998) argues that for-profit health plans enjoy monopsony power in the health care market. And while a monopsony can result in lower provider premiums, it may also reduce benefits as the providers reduce their spending on care. Pauly concludes that traditional antitrust policies have not addressed the problem of monopsony or monopoly power.

Woolhandler and Himmelstein (1995) view the growth in HMOs as a transformation that not only places physicians at financial risk because of cost-containment practices but also pressures physicians to withhold needed care. These financial arrangements encourage physicians to collaborate in risk selection by seeking to avoid patients who may have high care needs. Mechanic and Schlesinger (1996) have argued that the impact of managed care has had a negative impact on patients' trust of medical care and their physicians. The insurance industry and the managed care organizations hold great power and authority over the care provided to individuals and groups. They also have a great deal of influence over physicians, other health professionals, and health care workers. The current trends can have serious negative consequences for access and the health status of the population.

PUBLIC PAYERS

The two major public payers are Medicare and Medicaid, programs established in l965. Medicare pays for those individuals who are aged or disabled under Title 18 of the Social Security Act. Medicaid is generally provided to persons age 65 and over who are entitled to Social Security and to individuals who are disabled for a period of at least 24 months. There were 38 million people eligible in l997. Medicaid, under Title 19 of the Social Security Act, is funded jointly by federal, state, and local governments. It provides coverage to those persons who are eligible for Supplemental Security Income, for low-income individuals who are aged, blind, or disabled. It also pays for recipients and their children under the Aid to Families with Dependent Children, which is now the Temporary Assistance for Needy Families program. The program provided assistance to 36 million persons in l996 (Waid 1998). There are other public payers, such as the Veterans' Administration and local governments, that are not discussed here in detail.

Medicare. In l997, Medicare paid $214.6 billion for its 38.4 million aged and disabled enrollees (Levit et al. 1998). Medicare is the single largest public program, accounting for 20 percent of the nation's health expenditures. See Table 2. The growth rates in Medicare spending declined from 12.2 percent in l994 to 7.2 percent in l997. This decline was the result of slower price inflation and policy changes. Physician payments were limited, and fraud and abuse detection programs were increased. Limits on payments and aggressive abuse detection programs are credited with decreasing home health care spending (Levit et al. 1998). Future declines in nursing home spending are expected because of the adoption of a prospective payment system that was implemented in l998.

In the 1980s and l990s, Medicare spending growth had been much lower than private health insurance annual growth rates. But in l997, Medicare spending growth rates were 5.8 percent compared with 3.8 percent for private health insurance (Levit et al. 1998). Unfortunately, this growth in spending has contributed to public policy attention on reducing spending for the Medicare program.

And although spending for Medicare has increased, the program pays for only a limited amount of health insurance for those who are aged and disabled. Data from the 1997 Current Population Survey indicate that Medicare beneficiaries paid 67 percent of their total expenditures. Private insurance paid for only 10 percent of the bills, even though a large proportion of beneficiaries have private supplemental insurance. The remainder was paid out of pocket by beneficiaries (Carrasquillo et al. 1999).

When the Medicare, Medicaid, and other public programs are combined, their expenditures represent 46 percent of total national expenditures. Because this is such a large percentage of the total, the large increases in public spending per year contribute to the public policy focus on cost containment.

Medicaid. Medicaid is a program designed for the poor as a safety net. Medicaid spending was $159.9 billion in l997, or a 3.8 percent increase over the previous year (Levit et al. 1998). Other governmental expenditures for health care (federal, state, and local) totaled $132.7 billion. These public programs represented 27 percent of total U.S. expenditures. See Table 2. There was a rapid growth in Medicaid expenditures in the l988–1994 period, related to the growth in (1) the number of enrollees, (2) the spending per enrollee, and (3) increases in spending for hospitals with disproportionate shares of Medicaid patients. The spending slowed in the l994–1997 period because the number of Medicaid enrollees dropped during this period. The states also were allowed to have waivers to implement mandatory enrollment in managed care plans.

Congress passed welfare reform legislation in l996 that replaced the Aid to Families With Dependent Children (AFDC) with Temporary Assistance for Needy Families (TANF). This legislation broke the link between welfare and Medicaid eligibility, but states must continue to provide benefits to persons who met the eligibility requirements for Medicaid prior to the legislation. The law changed the definition of childhood disability under the Supplemental Security Income (SSI) program, curtailed benefits to legal immigrants, and banned Medicaid eligibility for qualified aliens admitted to the United States after the legislation. Some benefits were restored to legal immigrants under the Balanced Budget Act of l997 (Levit et al. 1998).

The passage of TANF means that welfare limits will gradually become effective in most states during l999, and millions of women will lose their eligibility for welfare benefits. Although states may continue to offer Medicaid eligibility to those individuals who go off welfare and are working in low-income jobs, many former welfare recipients are not informed of their eligibility for Medicaid services and thus have not retained their benefits. Those former welfare recipients who successfully obtain employment will generally be placed in relatively low-wage occupations that will probably not offer health insurance or job security. Overall, the legislation will result in a loss of health insurance for many women and children (Kuttner 1999).

In response to the health needs of children, Congress passed the Children's Health Insurance Program (CHIP) as a part of the l997 Balanced Budget Act. This program provided states with $24 billion over five years to expand coverage of Medicaid for children or to establish a new children's health insurance program. The new insurance program is complex and difficult for people to understand and gain access to. In l998, it was reported that the program was dramatically undersubscribed for a number of reasons across the country. California, for example, had enrolled only about 4 percent of the 580,000 children eligible for the program (Kuttner 1999). Some immigrants were reported to be concerned that enrollment of their children could affect their immigration status.

Several barriers restrict access to this program, particularly for immigrant populations. First, a large state bureaucracy and complicated application process have limited access to services. Second, eligibility coverage varies for children of different ages, so that one child may be covered while another is not in a single family. Many immigrant parents declined to participate in the program because they felt that providing health care for one child while not obtaining coverage for another was immoral (Altman 1998). Finally, eligibility for this program was initially tied to citizenship, which created access problems because, although their children were often documented and legal residents, undocumented immigrant parents declined to apply to the program for fear of deportation by immigration authorities (Korenbrot et al. 1999).

The Congressional Budget Office estimated that the Children's Health Insurance Program and the Medicaid expansions could extend coverage to 2 million of the 10.6 million uninsured children in l997 (Kuttner 1999). It is uncertain if more than a fraction of this total will ever receive coverage.

CONSUMER POWER

Consumers have seen a drop in their out-of-pocket spending for medical care over the past two decades. In l980, out-of-pocket spending was 24 percent of the total expenditures in contrast with 17 percent of spending in l997 (Levit et al. 1998). The total out-of-pocket spending was $187.6 billion in l997. Out-of-pocket costs include premiums for insurance, copayments, and deductibles for services use, and direct costs for services not covered by insurance, such as outpatient drugs and long-term care. The reason that out-of-pocket spending is lower is that some consumers are also paying a lower proportion of their income on premiums for health insurance. Managed care plans generally had more limited copayments and deductibles than indemnity insurance, and more individuals were enrolled in managed care (Ginsburg and Gabel 1998). Out-of-pocket costs did, however, go up in l997 because some HMOs were reportedly increasing their copayment and deductible charges (Levit et al. 1998). Consumer spending for drugs increased as well.

Some studies have reported that 61 percent of the population has employment related coverage (AHCPR 1998; Levit et al. 1998). These estimates group individuals who have public and private employment together and tend to count anyone with private insurance even if that insurance is not the primary payer of health care. New data from the 1997 Current Population Survey showed that only 43 percent of the population had their health insurance paid for by private-sector employers, 34 percent have publicly funded insurance, 7 percent purchased their own insurance, and 16 percent were uninsured (Carrasquillo et al. 1999). Employers also contribute to the Medicare Hospital Insurance Trust Fund through Social Security payroll taxes, but these account for only a small amount of the total expenditures.

Carrasquillo and colleagues (1999) point out that the role of employers in paying for insurance has been exaggerated and, as a consequence, employers have played a dominant role in public policy discussions related to health insurance. For example, President Clinton's national health insurance plan proposed to link health insurance to that paid for by private employers, even though employers are not the major payers of insurance. Employers played a role, along with many other interest groups, in defeating the Clinton plan because of fears that their costs would be increased and they would lose control over their work force (Navarro 1995). At the same time, employers are given an estimated $100 billion in tax subsidies in order to cover the costs of employer-sponsored health insurance (Reinhardt 1997). The $100 billion in subsidies could be eliminated if a national health insurance plan were paid directly by the government.

TYPES OF UTILIZATION AND EXPENDITURES

National expenditures for health care reveal how health resources are utilized and allocated to provide care in the United States. The patterns are not necessarily the way the public would allocate the resources. Instead, the expenditure patterns reflect historical decisions about how public and private dollars should be spent.

The patterns of expenditures changed rapidly in the latter half of the 1990s because of the growth in managed care. Managed care plans have instituted their own cost controls on the providers that they pay. The health industry has also been consolidating through a large number of mergers and acquisitions. These changes have resulted in fewer larger health care provider organizations.

Hospital Care. Hospital care has historically been and continues to be the largest component of health spending, accounting for 38 percent of total spending ($371 billion) in l997. See Table 1. The spending for inpatient care dropped considerably and marked increases occurred in outpatient services. This shift represents a restructuring to outpatient care, which is less costly. In the l990–l997 period, there was a 6-percent reduction in hospital admissions per capita and a 16-percent decline in inpatient days in community hospitals (American Healthcare Infosource 1998).

As result of the restructuring and downsizing, U.S. hospitals closed 88,000 beds (10 percent) and occupancy rates fell from 64.5 percent in l990 to 59.6 percent in l997 (Levit et al. 1998). Although utilization fell and price increases were controlled, hospitals were able to keep profit margins high, since almost all their revenue is from third-party insurance (consumers paid directly for only 3 percent of hospital costs).

Physician Services. Costs of physicians' services were $218 billion in l997, or 21 percent of total health services and supplies. Spending increases for physicians were associated with the increasing dominance of managed care organizations. The American Medical Association (AMA) reported that 92 percent of all physicians were in managed care, and these contracts accounted for 49 percent of their income (Levit et al. 1998). Physicians are also expected to pay for an increasing share of ancillary services under the capitated managed care contracts. Physicians have reacted to managed care cost controls by expanding the sizes of their practices to large group practices. These practices allow for greater leverage in negotiating contracts, economies of scale, and increases in capital. More research on the effects of the changing structure of medical practices on health care access, costs, and quality is needed.

Drugs and Medical Nondurables. Drugs and medical nondurables accounted for $108.9 billion in l997, or 10 percent of total national expenditures (Levit et al. 1998). Prescription drugs as a subset grew faster than other types of health care. These costs grew at 14.1 percent in l997 compared to 4.8 percent for health spending in 1989. Out-of-pocket drug costs were 51 percent of total payments in l989, later dropping to only 29 percent in l997, and the rest were paid by third parties. The switch to managed care has increased the amount of covered outpatient drugs with relatively low copayments (Levit et al. 1998). Thus, the growth in drug costs was associated with increases in the number of prescriptions—not price increases, as had been the case in the l980s (Levit et al. 1998).

The demand for drugs continues to increase. The Food and Drug Administration (FDA) approved fifty-three new drugs in l996 and thirty-nine in l997—both record highs. These approvals, along with increased advertising efforts by pharmaceutical firms, may lead to increases in demand. Drug makers reported spending $21 billion on research and development in l998, compared with $8 billion in l990 (Herrera 1999). They also increased spending on advertising to $1.3 billion in l998, having great successes with Viagra (for impotence), Claritin (for allergies), and Propecia (for baldness) (Herrera 1999).

Other Services. Nursing homes represented about 8 percent of total health service and supply expenditures in l997, and this is projected to decline to 7 percent in 2007. Other professional services were 4.6 percent of expenditures, home health care was 3 percent, dental services were 4.6 percent, and vision products and other durable medical equipment were only 1.4 percent in l996. These expenditures were expected to remain about the same percentage of the total over the next ten-year period (Smith et al. 1998). Governmental public health activities were estimated to be about $35.5 billion in l996, representing 3.4 percent of total U.S. health expenditures. Research expenditures were 1.6 percent of the total in l996, and construction was 1.4 percent of the total (Smith et al. 1998); these were also projected to remain the same over the next ten-year period.

LONG-TERM CARE IN THE UNITED STATES

The only segment of the U.S. population whose cost of long-term care is fully covered is made up of those individuals below the poverty threshold who are enrolled in the state-run, federally supported Medicaid plans (Harrington et al. 1991). Many persons of moderate incomes needing long-term care are unable to afford the costs of long-term care services, which can be as much as $50,000 per year for nursing home care. If individuals "spend down" to the poverty threshold, they can become Medicaid eligible as a last resort (Wiener 1996). This not only constitutes a hardship to the patient but creates dependence on federal and state assistance, which would be unnecessary if the entire population were insured, with premiums derived from sources other than the government.

In contrast, the nonpoor elderly enrolled in Medicare are entitled only to a limited number of skilled nursing care days (up to 100 days) if medically required following hospitalization. With some exceptions, the rest of the population must either pay for care out of pocket or purchase private long-term care insurance (Levit et al. 1997).

In 1996, national estimates for long-term care spending were $125.5 billion (Levit et al. 1997). Of the total expenditures, 30 percent ($38 billion) was for home health care (including hospitals and freestanding agencies) and 70 percent was for nursing home care ($87.5 billion; including hospital and freestanding facilities) (Levit et al. 1997). Home health expenditures are expected to double between l996 and 2007, while nursing homes are expected to grow by 50 percent during the same period (Smith et al. 1998).

Medicaid paid for 48 percent of all nursing home care and 14 percent of all home health care in 1996 (Levit et al. 1997). Medicare paid for 45 percent of home health care and 11 percent of nursing home care. Overall, the government paid 59 percent of home health costs and 61.5 percent of nursing home costs (Levit et al. 1997). Most of the burden for government spending is from general taxes used to pay for Medicaid and a combination of general taxes and payroll taxes that support the Medicare program.

Private health insurance paid only an estimated 5 percent of nursing home care and 10.6 percent of home health care expenditures in l996 (Levit et al. 1997). The remainder of the expenditures was paid directly out of pocket by those needing long-term care (31.5 percent for nursing home care and 19.5 percent for home health care) (Levit et al. 1997).

Private, voluntary long-term care insurance is not a viable approach to financing long-term care (Wiener 1994). Although private long-term care insurance has been available since the late l980s, only 4.5 million long-term care insurance policies had been sold by l994 (and not all of these policies were still in effect) (Cohen and Kumar 1997). Sold on an individual basis, private long-term care insurance is primarily attractive to persons whose health condition places them at high risk and makes them likely candidates for long-term care. Only about 10 to 20 percent of the elderly can afford to purchase long-term care insurance (Wiener 1994). Premiums for two policies purchased at age 65 were estimated to cost an average of $3,500 per year, which would be about 13 percent of a median elderly couple's income (Consumer Reports 1997, p. 46). These models determined that private long-term care insurance is unlikely to have a significant impact on public spending for long-term care, even though such insurance is expected to increase (Wiener and Illston 1994; Wiener et al. 1994).

A mandatory social insurance program for long-term care would have many advantages (Harrington et al. 1991). If everyone paid into the system, then individuals would have access to coverage when they are chronically ill or disabled without the humiliation of having to become poor (i.e., to "spend down") to receive services. The program might be more appealing if enrollees could have the advantage of paying in advance (prefunding) so that services would be available when needed. No stigma would be attached to receiving services, and such a program should have wide public support. The program could also reduce the access problems that are currently experienced by those who are in the Medicaid program. The financial risk would be spread across the entire population so that individual premium costs or taxes would be relatively low, in comparison to the costs of insurance purchased when individuals are older and at risk of needing long-term care.

Although many have argued that the United States cannot afford to adopt a public social insurance program, Germany mandated a social insurance program for long-term care in l995 (Geraedts et al. l999). This program was funded through a combination of public taxes and payroll taxes paid by employers and employees (.08 percent of wages for each or a total of 1.6 percent) (Geraedts et al. l999). During the first three years of this program, it maintained its financial solvency and expanded long-term care to the entire population who have disabilities. This example demonstrates that such public insurance programs, if paid for by the entire population, can ensure coverage at a reasonable cost. This approach avoids restricted access, the stigma, and the stress placed on those individuals needing long term care and their families who do not have sufficient funds to pay for them in the United States (Harrington et al. 1999).

THE UNINSURED AND THEIR HEALTH: MICRO-LEVEL ISSUES

An estimated 43 million individuals in the United States have no private health care insurance (Kuttner 1999; U.S. Bureau of the Census 1998b). These individuals have no Medicare, Medicaid, or other public insurance coverage. The percent of the population without insurance increased from 14.2 in l995, to 15.3 percent in l996, and to 16.1 percent in l997 (Kuttner 1999). This represents a steady increase in the number of uninsured since the l980s, when the Reagan administration first began cutbacks in public health expenditures. The U.S. Bureau of the Census (1998b) also estimated that about 71.5 million individuals lacked insurance for at least part of the year in l996.

The uninsured are primarily those who are poor and members of minority groups. Of those with incomes of less than $25,000, 24 percent had no health insurance compared with only 8 percent of those with incomes over $75,000 (Kuttner 1999). For those from minority groups, 50 percent of Hispanics and 37 percent of blacks had at least one month without insurance coverage compared to only 25 percent of whites (Kuttner 1999). Moreover, approximately one in three children had no health insurance for part of the year during l995 and l996 (Families USA 1997). The rate of children lacking insurance rose between l989 and l996 (U.S. Bureau of the Census 1998a). Lack of insurance for children was closely correlated with income. The group with the highest uninsured rate was young adults (38 percent, or twice the rate for other Americans), because many of them are dependent on their parents, who are working in low-wage jobs (AHCPR, 1998).

REASONS FOR THE GROWING NUMBERS OF UNINSURED

There are many reasons why individuals do not have health insurance, but the primary one is that private health insurance rates are too high relative to the incomes of those who are poor. The costs of insurance would represent about 26 to 40 percent of incomes of those who are poor, so many of these individuals are unable to afford coverage (Kuttner 1999). Another major problem is that 44 percent of those individuals who lost their jobs in the l993–1995 period also lost their health insurance coverage.

The Medicaid program does not provide health insurance to all those individuals living below the U.S. poverty rate. State Medicaid programs are allowed to establish their own eligibility standards for those on welfare (TANF) in order to limit the number who can be served. The Census Bureau reported that 49 percent of people who were fully employed but living below the poverty line had no Medicaid or private insurance, and this number increased to 52 percent in l996 (Kuttner 1999). Due to reductions in Medicaid coverage, the rate of poor children with insurance fell from 16.5 million to 15.5 million between l995 and l996 (U.S. Bureau of the Census 1998a). A U.S. General Accounting Office study (1996) found that nearly 3 million children who were eligible for Medicaid were not enrolled in the program because of inadequate outreach, fears by immigrants about immigration problems, and other barriers. Additionally, the combination of immigration reform and welfare reforms has produced a "chilling effect" on many immigrants and their health care providers (Korenbrot et al. 1999).

Another reason that health insurance coverage is eroding is the rising cost of health insurance premiums, especially for people who purchase individual premiums privately rather than through the job (Kuttner 1999). Kuttner (1999) also cites the trends toward temporary and part-time employment, where most workers do not have insurance. The rising costs of Medigap premiums for the elderly who are on Medicare is also a major problem. Finally, the trend away from community rating of insurance to individual and group ratings makes the costs substantially higher for individuals and some groups that have higher injury or illness rates. Employers are also reducing their supplemental health coverage for retirees (from 60 percent to only 40 percent of retirees in l995) (Kuttner 1999).

There has been a reduction in benefit coverage, particularly for pharmaceutical benefits. Many health plans are capping their outpatient drug benefits, although prescription drugs are the largest category of out-of-pocket costs for the elderly and the costs are increasing rapidly (Kuttner 1999). As noted earlier, drugs costs have been increasing rapidly (Levit et al. 1997) and many cannot afford to pay these costs. Many HMOs and insurance companies have been reducing drug benefits because of these cost increases. One study found that 84 percent of Medigap policies for the elderly had no drug coverage (McCormack et al. 1996). The result is that many elderly individuals are unable to afford drugs that are prescribed for them.

Another problem has been limited health insurance coverage for persons with mental health problems and mental illness. Mechanic and Rochefort (1992) document the deinstitutionalization of the mentally ill from public mental hospitals and the problems associated with inadequate mental health services. Inadequate housing and community health services, as well as the limits on mental health insurance coverage have all contributed to a general decline in mental health services.

CONSEQUENCES OF LIMITED ACCESS TO HEALTH CARE

The United States and South Africa continue to hold the unenviable distinction as the only two existing industrialized countries without national health insurance. Thus, when the high U.S. expenditures per capita are compared with those of other industrialized counties, the majority of these nations have their populations covered by insurance, while the United States is excluding 43 million individuals. The lack of health insurance leads to a number of serious problems in access to health care services.

There are many other health-related problems that plague the U.S. population in comparison to those of other nations. Among the top twenty-four industrialized nations, the United States ranks sixteenth in life expectancy for women and seventeenth for men, while it ranks twenty-first in infant mortality (Andrews 1995, p. 38). Although there were substantial reductions in childhood mortality in the United States between l950 and l993 (Singh and Yu 1996), the United States is well behind many other industrialized countries in childhood mortality (two to four times higher than Japan and Sweden) not only because of higher mortality from medical causes (e.g., heart disease) but also because of injuries and violence. Moreover, there are large differences across groups, with male children experiencing higher mortality rates and African-American children experiencing rates of more than twice that of white children (Singh and Yu 1996).

A number of studies have been conducted on the barriers to access related to the inability to pay for services (Berk et al. 1995; Braveman et al. 1989; Berk et al. 1995; Hafner-Eaton 1994; Mueller et al. l998; Weissman et al. 1991). Mueller and colleagues (1998) confirmed that lack of health insurance—regardless of race or ethnicity, or living in a rural environment—is the major determinant of the utilization of health care services.

Approximately 12 percent of all American families (12.8 million) experienced barriers to receiving needed health care services in 1996. Sixty percent of these families reported that the barrier to care was their inability to afford the care, and 20 percent cited insurance-related problems (AHCPR 1997). The barriers to health care included difficulties finding care, delays, or not receiving the care that was needed.

Nearly 46 million Americans were without a regular source of health care such as a doctor's office, clinic, or health center in 1996 (AHCPR 1997). Those persons without insurance were two to three times more likely to have no regular source of health care and two to three times more likely to have encountered barriers in receiving needed health care than those persons with health insurance (AHCPR 1997). Hispanics (30 percent), the uninsured under age 65 (38 percent), and young adults (34 percent) were more likely to lack a stable source of health care. Those individuals with a history of serious medical illness were twice as likely to be unable to obtain care. African Americans and Hispanics were also more likely to be unable to obtain care than whites and others (Himmelstein and Woolhandler 1995).

OTHER BARRIERS TO HEALTH CARE ACCESS

There are many other barriers to access to care. One is the lack of providers in areas close to people who need care. This is a problem for those persons living in rural areas as well as for those in central- or inner-city areas (Clarke et al. 1995). Weisgrau (1995) reviewed the myriad problems rural Americans confront in obtaining health care and mental health services due to the shortages of health care providers and the continued closures of rural hospitals. The stigma of obtaining mental health services in rural areas was also considered to be a problem.

Studies by Wennberg and colleagues (1989) have documented the wide geographic variations in health outcomes. These outcomes are considered to be, in part, related to variations in provider practices. A recent study by the Department of Veterans' Affairs found that even though the system predominantly serves low-income men, they found substantial geographic variation in service use for different diagnoses in hospitals and clinic use across the United States which they attribute to different practice styles (Ashton et al. 1999).

Another problem is cultural barriers to care. These include language differences between patients and providers and other communication problems. Escarce and colleagues (1993), Gornick and colleagues (1996), Friedman (1994), and Korenbrot and colleagues (1999) have identified these problems. Many studies have found persistent disparities across racial and ethnic groups in access to care (Escarce et al. 1993). Racial and ethnic minorities and rural residents were less likely to use physician services. This pattern was even stronger for rural Latinos and Asians (Mueller et al. 1998), particularly refugee populations. Mortality rates for black Americans are about 50 percent higher than for white Americans, and Native Americans also show very high death rates, especially among the young (Nickens l995). In part, the high mortality rates are related to lower socioeconomic status (SES) measured by occupation, income, and education attainment, which are generally lower for minorities. Mortality rates were more strongly related to SES in l986 than in l960. The substantial black and white differentials in infant mortality have been well documented (NCHS, 1994). High infant mortality rates for minorities are a function not only of poverty and low SES, but also of racial segregation in housing markets that are in close proximity to industrial pollution generators (incinerators, factories, etc.). These sources of pollution produce poor air quality in many communities of color and have a direct effect on infant mortality rates and adult morbidity (Hurley 1995).

In addition to poverty, low SES, and poor environmental quality, infant mortality rates of minorities are also influenced by continuing discrimination in society. The discrimination accumulates over time and results in perceived powerlessness, frustration, and negative self-images that in turn contribute to higher mortality rates. Discrimination against minority group members by health care providers may be a contributing factor that has not been eliminated. Using simulated video interviews of patients and physicians, a recent study documented that the race and sex of patients have an independent influence on physician decisions regarding how to manage care for chest pain. Women, blacks, and younger patients were less likely to be referred for diagnostic tests (cardiac catheterization), and the interaction effects were also significant (Schulman et al. 1999).

SOCIOLOGICAL MODELS OF ACCESS TO SERVICE USE

Much of the study of health care utilization and access has been developed using the theoretical model adopted by Andersen, Aday, and colleagues at the University of Chicago (Aday and Andersen 1974, 1991; Andersen 1995; Andersen and Aday 1978).

Andersen's (1995) original model was designed to predict or explain the use of health services. Three major components were developed for the model. First, predisposing characteristics were considered to include those that were endogenous to use including (1) demographics, such as age, gender, and race; (2) social structure, such as education, occupation, and ethnicity; and (3) health beliefs, including attitudes, values, and knowledge about health. Various critics have suggested (and Andersen 1995 agreed) that social networks, social interactions, and culture are all concepts that can be incorporated into the social structural factors. Genetic makeup might also be added to the list of predisposing factors. Second, enabling factors were considered to be those community and personal resources that may enable an individual or family to use health services. These include the supply of health personnel and organizations or facilities to provide care. Enabling factors also include income, health insurance, regular source of care, travel and waiting times, and other such factors. Social relationships, or "social support," are also important enabling factors according to Andersen (1995).

Finally, the original model included the need for health care services. "Need" included both the perceived need and the need as evaluated by health professionals. In the original model, the outcomes included the amount of physician, hospital, and other health services used. This model is one of the most widely used approaches by scholars, with hundreds of articles presenting the results of testing these factors. In Andersen's original model, access was described as potential access and realized access (actual use) of services. Access could be found to be equitable or inequitable, depending on value judgments. Inequitable access occurred when predisposing factors (such as race, ethnicity, or gender) or enabling factors (e.g., health insurance) determine who receives services rather than services determined by need. In later models, the health care system was incorporated, including national health policy, resources, and organizations. Outcomes of health services were revised to include consumer satisfaction about convenience, availability, financing, providers, and quality (Aday and Andersen 1974). The model was expanded further to include ethical criteria to measure equity, such as freedom of choice, equal treatment, and decent basic minimum care. Subsequently, the model has included understandings of effective and efficient access to services and improved health status as outcomes (Andersen 1995). Finally, Andersen sees the model as interactive and dynamic, with utilization and outcomes influencing predisposing, enabling, and need factors.

As a test of the importance of the behavioral model, a study in the late 1990s examined its use in the social science literature between 1975 and 1995. The study found 139 empirical research articles that used the model; of these articles, 45 percent included environmental variables and 51 percent included provider-related variables (Phillips et al. 1998). Extensive work has been done to examine health status as measures of health service outcomes. These include mortality, morbidity, well-being, and functioning. Outcomes can also include equity of service access. More elaborate conceptual models have been developed by the Institute of Medicine (Gold 1998; Millman 1993).

ISSUES RAISED BY TRENDS IN HEALTH CARE UTILIZATION AND EXPENDITURES

The macro-level and micro-level patterns of health care utilization and expenditure suggest that consumers, workers, the poor, people of color, the elderly, nurses, and even physicians are losing ground with respect to access to care and decision making within the health care system. Each year—indeed, every month—thousands more individuals join the ranks of the uninsured and thousands more remain underinsured (Kuttner 1999). At the same time, expenditures for the nation's health care system are expected to rise steeply in the first decade of the twenty-first century. While most policy makers and health policy analysts regard the American health care system as characterized by "private insurance," less than 50 percent of health insurance is in fact paid for by private employers—and these payments are subsidized by the government at a total of $100 billion (Carrasquillo et al. 1999). Thus, the role of employers in paying for insurance has been overstated and employers have played a prominent role in health policy discussions. And while the number of employers offering their employees health insurance has either remained unchanged or even decreased in recent years, the cost of these plans has been prohibitive for many workers, who then go without coverage. The result is a nationwide decline in persons with employer-sponsored health coverage (NCH Statistics 1994).

These trends raise questions about the role of the U.S. health care system in providing care to the nation's population. What becomes clear is that while the system is heavily supported by public funds, the benefits are accruing more and more to private corporations. Thus, we suggest that in order to understand changing patterns in health care utilization and expenditure, a political economy perspective is useful. A political economy perspective recognizes the enduring underlying conflicts among different interest groups (workers, purchasers, providers, doctors, insurers, nurses, the elderly) in gaining access to the myriad benefits a health care system provides. For example, many scholars argue that while the U.S. medical system is officially designed to provide the best possible care to any person who needs it and can pay for services, health care is actually only a secondary or tertiary function of this system (Navarro 1976; Relman 1980; Waitzkin 1983). Profit, above all else, is the "driving force" behind the health care system (Woolhandler and Himmelstein 1995).

The benefits and costs of health care in the United States are hotly contested, and stakeholders are engaged in a continuous struggle to gain access to the former and to externalize the latter. Doctors and nurses have organized unions; consumers have demanded a "patient's bill of rights" and a long list of other legislative reforms, producing a "managed care backlash"; health care providers and insurers have consolidated in a flurry of mergers and acquisitions in the 1990s; pharmaceutical and other health care firms have sought out markets around the globe; and managers of health care systems have slashed wages and jobs in an effort to contain costs (Andrews 1995). The results have been: (1) less regulation of the health care industry; (2) increasing profits for industry; (3) declining consumer satisfaction and less access to care and insurance under managed care; (4) loss of control and autonomy by doctors and nurses; (5) rising health care expenditures; and (6) no discernible improvement in the quality of health care. Our health care system is transforming rapidly without a coherent national approach to either the structure of the delivery of care or to the problem of rising expenditures. It is incumbent upon sociologists studying the health care system of this nation and others to provide a counterbalance to the currently dominant economic cost-benefit analytical approach to this topic. A sociological perspective offers a balanced understanding of the social basis on which the health care system operates and can, therefore, lead the way toward improvements in both theory and policy making in the areas of health and health care.

references

Aday, L. A., and R. M. Andersen 1974 "A Framework for the Study of Access to Medical Care." HealthServices Research 9(2):208–220.

——1991 "Equity to Access to Medical Care." Medical Care 19(12):4–27.

Agency for Health Care Policy and Research (AHCPR) 1997 Access to Health Care in America, 1996. Medical Expenditure Panel Survey, May (3):1–3.

——1998 Health Insurance Coverage In America – 1996. Medical Expenditure Panel Survey, May (4):1–3.

Altman, D. 1998 "The Kaiser Family Foundation: Our Mission." Paper presented to the Robert Wood Johnson Foundation Scholars in Health Policy Research Program, University of California, Berkeley.

American Healthcare InfoSource, Inc. 1998 HospitalStatistics. Chicago: American Hospital Association.

Andersen, R. M. 1995 "Revisiting the Behavioral Model and Access to Medical Care: Does it Matter?" Journalof Health and Social Behavior 36:1–10.

——, and L. A. Aday 1978 "Access to Medical Care in the U.S.: Realized and Potential." Medical Care 16(7):533–546.

Andrews, C. 1995 Profit Fever: The Drive to CorporatizeHealth Care and How to Stop it. Monroe, Maine: Common Courage Press.

Ashton, C. M., N. J. Petersen, J. Souchek, et al. 1999 "Geographic Variations in Utilization Rates in Veterans Affairs Hospitals and Clinics." New England Journal of Medicine. 340(1):32–39.

Berk, M. L., C. L. Schur, and J. C. Cantor 1995 "Ability to Obtain Health Care: Recent Estimates from the Robert Wood Johnson Foundation National Access to Care Survey." Health Affairs 14(3):139–146.

Blendon, R. J., J. Benson, K. Donelan, R. Leitman, H. Taylor, C. Koeck, and D. Gitterman 1995 "Who Has the Best Health Care System? A Second Look." Health Affairs 14(4):220–230.

Braveman, P., G. Oliva, M. A. Miller, R. Reiter, and S. Egerter 1989 "Adverse Outcomes and Lack of Health Insurance Among Newborns in an Eight-County Area of California, 1982 to 1986." New England Journal ofMedicine 321(8):508–513.

Carrasquillo, O., D. U. Himmelstein, S. Woolhandler, and D. H. Bor 1999 "A Reappraisal of Private Employers' Role in Providing Health Insurance." NewEngland Journal of Medicine 340(2):109–114.

Chassin, M. R., R. W. Galvin, and the National Roundtable on Health Care Quality 1998 "The Urgent Need to Improve Health Care Quality." Journal of the American Medical Association 280(11):1000–1005.

Christensen, S. 1998 "Medicare+Choice Provisions in the Balanced Budget Act of l997." Health Affairs 17(4):224–231.

Clarke, L. L.,C. A. Bono, M. K. Miller, and S. C. Malone 1995 "Prenatal Care Use in Nonmetropolitan and Metropolitan America: Racial/Ethnic Differences." Journal of Health Care for the Poor and Underserved 6(4):410–433.

Clement, D. G., S. M. Retchin, R. S. Borwn, and M. H. Stegall 1994 "Access and Outcomes of Elderly Patients Enrolled in Managed Care." Journal of theAmerican Medical Association 271(19):1487–1492.

Cohen, M. A., and A. K. N. Kumar 1997 "The Changing Face of Long-Term Care Insurance in l994: Profiles and Innovations in a Dynamic Market." Inquiry 34:50–61.

Consumer Reports 1997 "Long Term-Care Insurance: Special Report." Consumer Reports (October):36–50.

Davis, K., K. S. Collins, C. Schoen, and C. Morris 1995 "Choice Matters: Enrollees' Views of Their Health Plans." Health Affairs :99–112.

Escarce, J. J., K. R. Epstein, D. C. Colby, and J. S. Schwartz 1993 "Racial Differences in the Elderly's Use of Medical Procedures and Diagnostic Tests." American Journal of Public Health 83 (July):948–954.

Families USA 1997 One Out of Three: Kids Without HealthInsurance, 1995–1996. Washington, D.C.: Families USA.

Friedman, E. 1994 "Money Isn't Everything: Nonfinancial Barriers to Access." Journal of the American MedicalAssociation. 271(19):1535–1538.

Gallagher, L. 1999 "Health Care Services: The Big Money is in Small Towns." Forbes 163(1):182–183.

Geraedts, M., G. V. Heller, and C. Harrington 1999 "Germany's Long Term Care Insurance: Putting A Social Insurance Model into Practice." Manuscript submitted for publication.

Ginsburg, P. B., and J. R. Gabel 1998 "Tracking Health Care Costs: What's New in l998." Health Affairs 17(5):141–146.

Gold, M. 1998 "Beyond Coverage and Supply: Measuring Access to Healthcare in Today's Market." HealthServices Research 33(3):625–652.

Gornick, M. E., P. W. Eggers, T. W. Reilly, R. M. Mentnech, L. K. Fitterman, L. E. Kucken, and B. C. Vladeck 1996 "Effects of Race and Income on Mortality and Use of Services Among Medicare Beneficiaries." The New England Journal of Medicine 335(11):791–799.

Hafner-Eaton, C. 1994 "Patterns of Hospital and Physician Utilization Among the Uninsured." Journal ofHealth Care for the Poor and Underserved 5(4):297–315.

Harrington, C., C. Cassel, C. L. Estes, et al. 1991 "A National Long-Term Care Program for the United States." Journal of the American Medical Association 266(21):3023.

——, G. V. Heller, and M. Geraedts 1999 "Commentary on Germany's Long Term Care Insurance Model: Lessons for the United States." Manuscript submitted for publication.

Hellander, I., D. U. Himmelstein, S. Woolhandler, and S. Wolfe 1994 "Health Care Paper Chase, 1993: The Cost to the Nation, the States, and the District of Columbia." International Journal of Health Services 24(1):1–9.

Herrera, S. 1999 "Health Care Products." Forbes 163(1):180–181.

Himmelstein, D. U., and S. Woolhandler 1995 "Care Denied: US Residents Who Are Unable to Obtain Needed Medical Services." American Journal of PublicHealth 85(3)341–344.

——1986 "Cost Without Benefit: Administrative Waste in the U.S." New England Journal of Medicine 314:440–441.

Hurley, A. 1995 Environmental Inequalities: Race, Class,and Industrial Decline in Gary, Indiana 1945–1980. Chapel Hill: University of North Carolina Press.

Iglehart, J. K. 1995 "Medicaid and Managed Care." TheNew England Journal of Medicine 332(25):1727–1731.

Jensen, G. A., M. A. Morrisey, S. Gaffney, and D. Liston 1997 "The New Dominance of Managed Care: Insurance Trends in the 1990s." Health Affairs 16:125–136.

Kaiser Commission on the Future of Medicaid 1995 Medicaid and Managed Care: Lessons from the Literature. Menlo Park, Calif.: Henry J. Kaiser Family Foundation.

Kassirer, J. P. 1996 "Mergers and Acquisitions—Who Benefits? Who Loses?" The New England Journal ofMedicine 334(11):722–723.

Kenkel, P. J. 1993 "Provider Based Managed-Care Plans Continue Growth Trend." Modern Healthcare (May 19):26–34.

Korenbrot, C., L. S. Park, R. Sarnoff, J. Greene, and C. Bender 1999 Impact of Recent Welfare and ImmigrationReforms on Low-Income Pregnant Women and TheirHealth Care Providers. Unpublished report to the California Endowment. Institute for Health Policy Studies, University of California at San Francisco.

Kuttner, R. 1999 "The American Health Care System: Health Insurance Coverage." The New England Journal of Medicine 340(2):163–168.

Levit, K. R., C. Cowan, B. R. Braden, B.R., et al. 1998 "National Health Expenditures, 1997: More Slow Growth." Health Affairs 17(5):99–110.

——, H. C. Lazenby, B. R. Braden, et al. 1997 "National Health Expenditures, 1996." Health Care Financing Review 19(1):161–200.

Light, D. W. 1992 "The Practice and Ethics of Risk-Rated Health Insurance." Journal of the AmericanMedical Association 267(18):2503–2508.

McCormack, L. A., P. D. Fox, T. Rice, and M. L. Graham 1996 "Medigap Reform Legislation of l990: Have the Objectives Been Met?" Health Care Financing Review 18(1):157–174.

Mechanic, D., and D. A. Rochefort 1992 "A Policy of Inclusion for the Mentally Ill." Health Affairs 11(1):128–150.

——, and M. Schlesinger 1996 "The Impact of Managed Care on Patients' Trust in Medical Care and Their Physicians." Journal of the American MedicalAssociation 275(21):1693–1697.

Miller, R. H. 1998 "Healthcare Organizational Change: Implications for Access to Care and Its Measurement." Health Services Research 33(3):653–680.

——, and H. S. Luft 1994 "Managed Care Plan Performance Since l980: A Literature Analysis." Journalof the American Medical Association 271(19):1512–1519.

——, and H. S. Luft 1997 "Does Managed Care Lead to Better or Worse Quality of Care?" New EnglandJournal of Medicine 16(5):1–18.

Millman, M. (ed.) 1993 Access to Health Care in America.Institute of Medicine Committee on Monitoring Access toPersonal Health Services. Washington, D.C.: National Academy Press.

Mueller, K. J., K. Patil, and E. Boilesen 1998 "The Role of Uninsurance and Race in Healthcare Utilization by Rural Minorities." Health Services Research 33(3):597–610.

National Center for Health Statistics (NCHS) 1994 Health,United States, 1993. Hyattsville, Md.: U.S. Public Health Service.

Navarro, V. 1995 "Why Congress Did Not Enact Health Care Reform." Journal of Health Politics, Policy andLaw 20:455–461.

——1976 Medicine Under Capitalism. New York: Prodist.

Nelson, L., R. Brown, M. Gold, A. B. Ciemnecki, and E. Docteur 1997 "Access to Care in Medicare HMOs, 1996." Health Affairs 16(2):148–156.

Nickens, H. W. 1995 "The Role of Race/Ethnicity and Social Class in Minority Health Status." Health Services Research. 30(31):151–162.

Pauly, M. V. 1998 "Managed Care, Market Power, and Monopsony." Health Services Research 33(5):1439–1460.

Pear, R. 1998 "HMOs are Retreating from Medicare, Citing High Costs." New York Times (October 2):A20.

Phillips, K. A., K. R. Morrison, R. Andersen, and L. A. Aday 1998 "Understanding the Context of Healthcare Utilization: Assessing Environmental and Provider-Related Variables in the Behavioral Model of Utilization." Health Services Research 33(3):571–596.

Reinhardt, U. E. 1997 "Wanted: A Clearly Articulated Social Ethic for American Health Care." Journal of theAmerican Medical Association 278:1446–1447.

Relman, A. S. 1980 "The New Medical-Industrial Complex." New England Journal of Medicine 303:963–970.

Schulman, K. A., J. A. Berlin, W. Harless, J. F. Kerner, et al. 1999 "The Effect of Race and Sex on Physicians' Recommendations for Cardiac Catheterization." TheNew England Journal of Medicine 340(8):618–626.

Shaughnessy, P. W., R. E. Schlenker, and D. F. Hittle 1994 "Home Health Care Outcomes Under Capitated and Fee-For-Service Payment." Health Care FinancingReview 16(1):187–221.

Singh, G. K., and S. M. Yu 1996 "US Childhood Mortality, 1950 Through 1993: Trends and Socioeconomic Differentials." American Journal of Public Health 86(4):505–512.

Smith, S., M. Freeland, S. Heffler, D. McKusick, and the Health Expenditures Projection Team 1998 "The Next Ten Years of Health Spending: What Does the Future Hold?" Health Affairs 17(5):128–140.

Srinivasan, S., L. Levitt, and J. Lundy 1998 "Wall Street's Love Affair with Health Care." Health Affairs 17(4):158–164.

U.S Bureau of the Census 1998a Census Brief: ChildrenWithout Health Insurance. Washington, D.C.: U.S. Government Printing Office.

——1998b Health Insurance Coverage, 1997. Washington, D.C.: U.S. Government Printing Office.

U.S. General Accounting Office 1996 Health Insurancefor Children: Private Insurance Coverage Continues toDeteriorate. GAO/HEHS 96-129. Washington, D.C.: U.S. General Accounting.

Waid, M. O., U.S. Health Care Financing Administration (USHCFA) 1998 Overview of the Medicare and Medicaid Programs. Health Care Financing Review 1998 Statistical Supplement.

Waitzkin, H. 1983 The Second Sickness: Contradictions ofCapitalist Health Care. New York: Free Press.

Weisgrau, S. 1995 "Issues in Rural Health: Access, Hospitals, and Reform." Health Care Financing Review 17(1):1–14.

Weissman, J. S., R. Stern, S. L. Fielding, and A. M. Epstein 1991 "Delayed Access to Health Care: Risk Factors, Reasons, and Consequences." Annals of Internal Medicine 114(4):325–331.

Welch, W. P. 1996 "Growth in HMO Share of the Medicare Market, 1989–1995." Health Affairs 15(3):201–214.

Wennberg, J. E., J. L. Freeman, R. M. Shelton, and T. A. Bubolz 1989 "Hospital Use and Mortality Among Medicare Beneficiaries in Boston and New Haven." New England Journal of Medicine 321:1168–1173.

Wiener, J. M. 1996 "Can Medicaid Long-Term Care Expenditures for the Elderly Be Reduced?" The Gerontologist 36(6):800–811.

——, and L. H. Illston 1994 "How to Share the Burden: Long-Term Care Reform in the l990s." TheBrookings Review 12(2):17–21.

——, and R. J. Hanley 1994 Sharing the Burden: Strategies for Public and Private Long-Term Care Insurance. Washington, D.C.: Brookings Institution.

Woolhandler, S., and D. U. Himmelstein 1995 "Extreme Risk—The New Corporate Proposition for Physicians." NewEnglandJournalofMedicine 333(25):1706–1708.

CHARLENE HARRINGTON

DAVID N. PELLOW