NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.

Institute of Medicine (US) Committee on Health Care for Homeless People. Homelessness, Health, and Human Needs. Washington (DC): National Academies Press (US); 1988.

Cover of Homelessness, Health, and Human Needs

Homelessness, Health, and Human Needs.

Show details

4Access to Health Care Services for Homeless People


In a nation abundantly endowed with hospitals, physicians, and advanced health care technology, millions of Americans have difficulty obtaining health care. Problems of access to care especially affect the poor, members of minority groups, and residents of most inner cities and many rural areas. Homeless people often face additional obstacles to receiving health care services.

This chapter reviews factors that directly determine how homeless people do or do not obtain physical and mental health services, first, as a part of the larger population of the poor and then, more specifically, as homeless people. It begins by discussing the following questions:

  • Why do poor people use health care services to a lesser extent than non-poor people?
  • What are the barriers to health care specifically faced by homeless people?
  • What is the role of public and private health insurance and the impact of Medicaid in providing financial access?

The chapter then discusses the role of the traditional system of health care for the indigent (including the National Health Service Corps) and examines its role in filling the gaps resulting from the failure of Medicaid to provide access. The chapter then reviews two specialized health care systems: one to treat mental illness and the other to provide services to veterans.

Health Care of the Indigent

A right to health care has never been established in the United States either by statute or case law, but a broad base of public opinion generally appears to support the notion that people should be able to get medical attention when they need it, regardless of economic or social circumstances (New York Times, December 1, 1987). To quote from the 1983 report of the President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research:

The Commission concludes that society has an ethical obligation to ensure equitable access to health care for all. This obligation rests on the special importance of health care: its role in relieving suffering, preventing premature death, restoring functioning, increasing opportunity, providing information about an individual's condition, and giving evidence of mutual empathy and compassion. Furthermore, although life-style and the environment can affect health status, differences in the need for health care are for the most part undeserved and not within the individual's control.

The President's commission also stated that ''equitable access to health care requires that all citizens be able to secure an adequate level of care without excessive burdens" and that "efforts to contain rising health costs are important but should not focus on limiting the attainment of equitable access for the least well served portion of the public" (President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research, 1983).

Reality, however, has not matched that aspiration. For example, in 1986, the Robert Wood Johnson Foundation funded a study of a random sample of the general public to determine to what extent people have access to health care in the United States (Robert Wood Johnson Foundation, 1987). Along the lines of a previous study conducted in 1982 (Aday et al., 1984), the surveyors interviewed 10,130 people (adults and adults responding on behalf of their minor children), primarily by telephone contact. A separate sample of 300 households without telephones was interviewed to increase the representative nature of the sample. Eighteen percent of Americans surveyed were unable to identify a usual source of medical care—a question that health services researchers consider a benchmark for access to care. It should be noted that 18 percent of the population is equal to 40 million people. The proportion of the population without a usual source of care had increased two-thirds since the survey conducted in 1982.

A somewhat similar survey, the National Health Interview Survey (NHIS), is conducted on a continuous basis by personnel of the U.S. Bureau of the Census. Reports of data drawn from this survey are published at various times by the National Center for Health Statistics, a division of the Public Health Service of the U.S. Department of Health and Human Services. The method used in this survey—personal interviews based on a random sample of households—is generally considered by statisticians to be more reliable than telephone interviews, especially in obtaining a valid sample of the general population.

Comparisons with the data obtained in the Robert Wood Johnson Foundation survey (1987) must be limited because the NHIS survey does not estimate how many people lack a "usual source of health care." However, both surveys asked the respondents if they had seen a physician at any point in the year prior to the interview. The Johnson survey reported that 33 percent of the respondents answered in the negative (up from 19 percent in the 1982 survey). NHIS reported that 24 percent of the respondents answered in the negative in their survey; this was the same percentage as that in each of the previous 5 years (Kovar, 1987).

There are many reasons why people might not use medical care; the most obvious is that some do not need it. In the aggregate, Americans were probably healthier in 1986 than they were in 1982, and young adults continue to use medical care less frequently than do older people. There are, however, certain circumstances under which the need for medical service is broadly supported both by professional consensus and popular opinion, and in those circumstances the Johnson survey found some disheartening results. Seventeen percent of persons with identified chronic illnesses (e.g., diabetes or heart disease) that require regular medical supervision did not see a physician in the previous year; 20 percent of persons with hypertension had not had their blood pressure checked in a year; and 15 percent of pregnant women did not receive prenatal care during their first trimester.

People in fair or poor health with incomes below 150 percent of the poverty level visited physicians 27 percent less often than did non-poor people in poor or fair health; blacks in fair or poor health had one-third fewer physician visits than did non-Hispanic whites. Poor children saw physicians in 1986 as often as non-poor children did, but poor children were three times as likely to be reported as being in poor or fair health. Poor people were somewhat more likely to be hospitalized in 1986 than non-poor people, but they were substantially more likely to have serious health problems of the sort that ordinarily lead to hospitalization. Controlling for degree of illness, the poor received substantially less hospital care, physicians' services, and dental care than the non-poor (Robert Wood Johnson Foundation, 1987).

Data from NHIS are not adjusted by severity of illness. Even so, NHIS found that poor children (under age 17) had fewer physician visits than non-poor children. Poor adults (between ages 17 and 64) and the poor elderly (over age 64) had a higher average number of such visits per year than the non-poor in the same age ranges; however, it has long been established that the poor have a significantly higher burden of illness than the non-poor (National Center for Health Statistics, 1988).

Ethnicity, location, education, and social circumstances contribute to difficulties in obtaining needed medical care, but the primary reason for access problems among the poor and near-poor is financial. Obviously, the inability to pay for care is a condition that also applies to the homeless poor (Robertson and Cousineau, 1986). Nine percent of low-income people reported to the Johnson surveyors instances in which they had been deterred from seeking medical care for financial reasons or had sought care and had been turned away (Freeman et al., 1987). Predictably, the greatest difficulties in obtaining medical care are experienced by low-income people without health insurance, either public or private. As shown in Figure 4-1, the proportion of those deterred from care by financial considerations is twice as great (13 percent) among the uninsured as that among the population as a whole (6 percent).

Figure 4-1. Percentage of Americans not receiving health care for economic reasons in 1986.

Figure 4-1

Percentage of Americans not receiving health care for economic reasons in 1986. SOURCE: Kovar (1987).

The most likely explanation for the Johnson survey's finding of reduced access to health care compared with that 5 years ago, especially for low-income Americans, is the substantial increase in the number of people without health insurance. Although the numbers are subject to interpretation and are not derived from entirely consistent data sources, the number of Americans without health insurance at any given time has increased from roughly 25 million in 1977 to perhaps as many as 37 million or 38 million today. Slightly more than half of the uninsured live in households with incomes that are less than 150 percent of the poverty level. More than half of the uninsured live in households that have at least one employed person (Sulvetta and Swartz, 1986). In addition, as Table 4-1 indicates, those people without health insurance coverage are concentrated primarily in the younger segment of the U.S. Population and among minorities, two subpopulations that are disproportionately represented among the homeless.

TABLE 4-1. Health Insurance Coverage, Fourth Quarter, 1985 (numbers in thousands).


Health Insurance Coverage, Fourth Quarter, 1985 (numbers in thousands).

Here, the NHIS data tend to support the Johnson survey, at least concerning the extent of the problem of a lack of health insurance among the poor. NHIS reported that 28 percent of families with an annual income below $10,000 lacked any form of health insurance, even Medicaid. Because NHIS found that virtually all respondents in their survey who were over age 65 had some coverage (usually Medicare, frequently with some supplemental coverage), the more accurate figure for those with incomes below $10,000 per year is that for the poor population under age 65. When the elderly are removed from the calculation, the proportion of the poor without any coverage rises to 32 percent for those with incomes under $5,000 per year and to 37 percent for those with incomes between $5,000 and $9,999 per year. According to NHIS, among those under age 65, the groups most likely to lack coverage (irrespective of income) are those who are unemployed (38 percent), the poor (35 percent), those with less than 12 years of education (26 percent), those between the ages of 18 and 24 (25 percent), and blacks and others (20 percent each). As to why people lack coverage, 63 percent of the respondents to the NHIS survey who did not have coverage identified the inability to afford such insurance as their primary reason (U.S. Department of Health and Human Services, 1987).

For the proportion of uninsured people to be so great and to be increasing appears to be unprecedented in recent history. Although the number of uninsured people increased with the growth of unemployment during the recession of 1981-1982 (as was the case in earlier recessions), it has continued to increase during the period of economic growth that has largely prevailed since then. Beginning in the 1930s, economic growth in the United States has invariably been accompanied by expanded and improved health insurance coverage (along with other employment-based fringe benefits), but that pattern no longer holds. As Table 4-2 indicates, the percentage of uninsured has steadily increased since 1980 (there was a slight decrease of 0.2 percent in 1986).

TABLE 4-2. Nonelderly Population Without Health Insurance, 1980-1986.


Nonelderly Population Without Health Insurance, 1980-1986.

Over the past several decades, health insurance has been tied primarily to employment for most nonelderly Americans or to income support programs through Medicaid for those outside the labor market. The growth in the number of uninsured people during the past decade has occurred largely among the employed, and may be attributable to changes in the labor market. Employment in manufacturing industries, which have traditionally provided the most generous insurance benefits, has fallen, and the most dramatic employment increases have occurred in service industries in which employee benefits have historically been and continue to be less generous. In addition, although the unemployment rate has fallen in the past several years, the labor force has expanded dramatically, so that the number of unemployed people remains quite sizable.

Among low-income people, access to health care is often difficult, even for those who have Medicaid coverage or private insurance. Access to health care is still more difficult for people without any form of health insurance. Extending health insurance to those not presently covered is a critical issue in terms of preventing homelessness, especially as it may result from the financial consequences of illness or injury. As indicated earlier, one principal barrier to care is financial, but other barriers include an undersupply of services in many poor communities, provider hostility or resistance, cultural differences, and transportation.

This last barrier—transportation—is especially crucial for the rural poor. Transportation to and from health care facilities in rural areas may be largely unavailable, and clinics may be located many miles away. During the site visits to rural areas in Alabama, Minnesota, Mississippi, and North Dakota, members of the committee heard references to poor people in need of health care paying a friend or neighbor to transport them to a health care program, with no guarantee of a return trip home. Health care could literally be out of reach for the rural poor and the rural homeless.

Both the health care system for the indigent and the Medicaid system impose considerable noneconomic costs on their users as a result of resource scarcity, governmental suspicion of the beneficiaries of public programs, or both. Waiting times are often long at hospital clinics or income maintenance eligibility centers; physical amenities are generally few; and the hours of operation often appear to have been established to suit the convenience of the facilities' employees rather than that of its clients. These problems are encountered by many low-income people, but they have especially severe consequences for the homeless.


In the United States, the major program for the provision of health insurance for low-income people now covers a smaller fraction of the poor than it did at any time in the past decade. Aggregate Medicaid enrollment has not grown, but the number of poor people has increased by roughly 40 percent (U.S. Department of Health and Human Services, 1985). To understand the importance and limitations of Medicaid as a vehicle for providing access to health care for low-income people, including the homeless, it is necessary to penetrate a thicket of Medicaid practices, beginning with two of Medicaid's fundamental precepts: eligibility and benefits.


Medicaid was erected on the structures of the preexisting income maintenance programs provided under the Social Security Act, specifically, Aid to Families with Dependent Children (AFDC) and Aid to the Blind, the Disabled, and the Old-Aged, all of which (except AFDC) were federalized and subsumed under Supplemental Security Income (SSI) in 1972, (Stevens and Stevens, 1974). Eligibility for either AFDC or SSI automatically confers Medicaid eligibility on the recipient, except in a few complicated instances in certain states. In addition, states may provide Medicaid for "categorically related people," those in households that would be eligible for AFDC or SSI except that their incomes exceed eligibility standards by no more than one-third. Categorically related people may also be covered, at state option, if they are "medically needy," meaning that their total income minus their medical expenses is less than a set level determined by the state. That level ranges from the maximum standard of income eligibility up to 133 1/3 percent of that standard.

As with income maintenance programs, Medicaid is a federal-state program financed in part (about 54 percent) by the federal government, but administered by the states under federal rules and supervision. The latitude for state discretion is extremely wide and the result is extraordinary diversity. It is often suggested that Medicaid is best understood as 54 separate programs (the 50 states plus the District of Columbia, Puerto Rico, the Virgin Islands, and Guam), rather than a single, unitary program. Thus, individuals with incomes as high as $414 a month, before medical expenses, are eligible for Medicaid in Wisconsin, although an income in excess of $117 a month would render individuals ineligible in Tennessee. For a family of four in Alabama, $147 a month is the cutoff point for eligibility, but a family of four in California is "medically needy" at an income level of $801 per month (U.S. Department of Health and Human Services, 1985). Table 4-3 illustrates the wide diversity from state to state of the threshold of eligibility based on annual income (and percentage of the official poverty level) for the three major programs (AFDC, medically needy, and SSI) that enable a family or individual to be eligible for Medicaid. Interstate variation extends to eligibility categories as well as income levels. A particularly dramatic example is the contrast between the 27 states that provide AFDC (and thus, automatically, Medicaid) to two-parent households with children in which the principal wage earner is unemployed and the 23 states in which such households are ineligible for Medicaid (U.S. Congress, House, Committee on Ways and Means, 1987).

TABLE 4-3. Annualized Medicaid Eligibility Thresholds (AFDC, Medically Needy, SSI) as a Percentage of Poverty, January 1987.


Annualized Medicaid Eligibility Thresholds (AFDC, Medically Needy, SSI) as a Percentage of Poverty, January 1987.

Perhaps of still greater concern, in terms of the problems of homeless people in obtaining Medicaid, is the extraordinary confusion surrounding Medicaid eligibility for SSI recipients or those who would be Medicaid recipients were it not for differences between federal and state eligibility standards. For example, 30 states have eligibility requirements for Medicaid that are identical to the federal SSI eligibility requirements and have a single application process; 6 states have a single eligibility requirement but separate application processes; and 14 states have separate application processes and separate—and more restrictive—eligibility standards for Medicaid (U.S. Congress, House, Committee on Ways and Means, 1987).

To complicate matters still further, 29 states operate state-only programs, which cover individuals who are not eligible under federal law, entirely at state (or combined state and local) expense. Data on state-only programs are even more fragmentary and limited than data on other aspects of Medicaid, but it appears that state-only programs largely provide benefits for individuals. These programs are directed primarily to recipients of general assistance or other non-federally supported income maintenance programs and provide relatively comprehensive benefits to significant numbers of people only in New York, Illinois, and Maryland (U.S. Department of Health and Human Services, 1985).

The net result of Medicaid eligibility practices is that only slightly over half of all people with incomes below 150 percent of the poverty level are covered by Medicaid at any one time. Among the clients of the Johnson-Pew Health Care for the Homeless (HCH) projects who were seen more than once and for whom benefit status is known (N = 10,555), only 21 percent were participating in Medicaid (with another 5 percent participating in Medicare and 6 percent receiving some Veterans Administration benefits) (Wright and Weber, 1987). Medicaid enrollment has remained roughly constant nationwide since the late 1970s. Because the number of poor persons has increased since then, the proportion covered by Medicaid has fallen. Moreover, although the total Medicaid population has remained roughly constant, an increasing proportion of recipients consists of chronically ill elderly people in nursing homes (the largest single Medicaid expenditure) or other physically disabled, chronically impaired people (U.S. Congress, House, Committee on Ways and Means, 1987).

These aggregate figures mask enormous interstate variation. In 1982, the range of Medicaid recipients as a percentage of people below the poverty level was from 102 percent in Hawaii to 17 percent in South Dakota. In the more populous states, the percentages ranged from 72 percent in Michigan to 47 percent in Ohio to 20 percent in Texas (U.S. Department of Health and Human Services, 1985). In both 1985 and 1986, Congress took some limited steps to reverse the trend toward declining

Medicaid enrollments for pregnant women and poor children. It did this by mandating coverage of certain classes of first-time pregnant women and infants and by permitting states to decouple eligibility for Medicaid from receipt of cash assistance, thus permitting higher eligibility standards for Medicaid than for cash assistance. Data are not yet available, but only a minority of states have undertaken substantial expansions of eligibility standards, and the net increase in enrollments appears to be relatively modest (U.S. Congress, House, Committee on Ways and Means, 1987).


A substantial body of data demonstrates that Medicaid recipients have better access to health care than do low-income people without Medicaid or other forms of health insurance, but even for those who are eligible for Medicaid, the program has limits (Rogers et al., 1982). There is considerable latitude left to the states to define covered services and to determine how those services will be paid for.

Under federal law, every state that participates in the Medicaid program must cover inpatient hospital care, hospital outpatient services, physician services, rural health clinic services, other laboratory and x-ray services, skilled nursing facilities for adults, family planning, home health care services, and nurse-midwife services, although states may impose a number of restrictions or limitations on the use of these services. Almost all participating states also provide at least some coverage for other items, such as prescription drugs, although here the patchwork of limitations and restrictions becomes more formidable (U.S. Department of Health and Human Services, 1985).

More important than coverage limitations are payment restrictions. Before 1982, states were essentially required to pay Medicare-defined costs for inpatient hospital care, and they are still required to pay cost-related rates for inpatient hospitalization and nursing homes. However, they can pay essentially whatever they choose for other services, of which the most important are physicians' services. Thus, in 1984 the Medicaid program in Connecticut paid $6.75 for a routine office visit to a specialist, New York paid $9.00, Illinois paid $10.50, California paid $12, and a relatively generous state like Wisconsin paid $15 (U.S. Department of Health and Human Services, 1985). As a result, in most of the nation, only some physicians routinely accept Medicaid patients; conversely, the availability of Medicaid coverage has done little to stem the exodus of private practitioners and other providers of health care from low-income communities. Access to care, even for those with Medicaid coverage, is particularly problematic for those needing obstetrical care, elective surgical procedures, and certain other types of specialty care, including psychiatry (Holahan, 1984). The net effect is that even when people are covered by Medicaid, they are effectively shut out of the traditional private practice fee-for-service system and must rely on the hospital system (especially the public hospital) for physician's services (Davis et al., 1981). As discussed later in this chapter, that system is already heavily involved in providing uncompensated care to the uninsured.

Medicaid and the Homeless

For homeless people, as with other low-income people, the availability of Medicaid is erratic and at least partially a function of circumstances. Single-parent families with no income other than cash assistance are generally eligible for Medicaid coverage; single individuals without children or certified disabilities generally are not. For other low-income people, it depends on such matters as the political jurisdiction in which they reside, the size and sources of their income and assets, and their skill in negotiating the bureaucratic process. As for other poor people, Medicaid coverage can be a major benefit to the homeless in removing a primary barrier to receiving health care, but it rarely solves all access problems. Many providers of health care services, especially institutions such as hospitals and free-standing health centers, certainly are more willing to serve the poor with Medicaid coverage than those without any financial resources. Yet there are shortages of health care providers (especially private physicians) in many communities occupied by high concentrations of low-income people. In addition, Medicaid generally does little to overcome such barriers as transportation cost or difficulty, cultural differences between health care providers and potential clients, or inconvenient hours or child care arrangements (Davis et al., 1981). Furthermore, when a homeless person qualifies for Medicaid as a result of receiving SSI benefits, that same homeless person who resides in a public shelter for more than 3 months in any calendar year loses the SSI eligibility and, with that, eligibility for Medicaid. The expansion of Medicaid, and especially its decoupling from other entitlement programs, could be an important step toward improving the access of low-income Americans to health care services. In many instances, however, including that of homeless people, it would probably not be a sufficient step. The homeless encounter additional barriers to the receipt of health care services that other poor people often do not.

The Indigent Care System

Many people get care without having Medicaid, private insurance, or personal financial resources. They received medical care even before

Medicaid was enacted. Separate from the Medicaid program, but closely and increasingly tied to it, there exists in some communities a network of health care services for low-income people that is generally less than fully adequate but that provides care for some people. This system draws on a range of resources separate from Medicaid financing.


The most important component of this system is the several hundred public hospitals located in inner city and low-income rural areas. They are supplemented by perhaps a roughly equivalent number of voluntary (private, not-for-profit) hospitals that, in terms of location, types of patients served, and financial resources, often closely resemble public hospitals. Many of these institutions, both public and private, are teaching hospitals that have been engaged since their inception in service to the uninsured poor, both as a reflection of a general mission and as a way of ensuring an adequate supply of patients for teaching purposes.

Using 1980 data, researchers identified 1,800 community hospitals—approximately 30 percent of the nation's total—that provide 70 percent of the hospital care to the poor, including those covered by Medicaid and those for whom hospitals receive no reimbursement at all. Fewer than 1,000 hospitals provide almost half of all care to the poor. A total of 38 percent of high-volume hospitals, as compared with 28 percent of other hospitals, are public; and 13 percent are private, not-for-profit teaching hospitals. The role of hospitals in caring for poor people is indeed highly variable, but it is not inconsistent with the data to characterize 400 to 500 urban hospitals, and a similar number of rural hospitals, as the backbone of the health care system for the indigent (Feder et al., 1984).

Altogether, the nation's hospitals provide something over $6 billion a year in bad debt and charity care, primarily to low-income people, in addition to roughly $16 billion in services to Medicaid recipients (American Hospital Association, 1986). Because the numbers of Medicaid-covered and non-Medicaid-covered poor people are roughly equal, these figures might be taken to imply that the Medicaid-covered poor receive two and a half times as many hospital services per capita as the non-Medicaid-covered poor. In fact, the dynamics of the eligibility criteria for the medically needy are such that people who need more expensive services are more likely to become eligible for Medicaid. However, the willingness of many hospitals to serve uncovered individuals does not provide the equivalent access to care as that which would exist under Medicaid.

Much too little is known about the characteristics of uncompensated care provided by hospitals, or the characteristics of patients who receive such care. However, several generalizations are possible. In many big cities, a substantial proportion of unreimbursed services is provided to outpatients. Hospital outpatient departments and emergency rooms provide routine medical care for thousands of poor people. Reliance on such services as principal providers of primary and specialized care has long been justly criticized in terms of its impersonality, lack of continuity, and expense. Efforts have been made in many communities to divert outpatient services from traditional clinics and emergency rooms to group practice arrangements, off-site clinics, or prepaid health plans. However, the entrenched system of outpatient and emergency room care is resilient.

Uncompensated inpatient services tend to be highly concentrated in a few diagnoses and types of cases, especially maternity and trauma services (Sloan and Valvona, 1986). Relatively little uncompensated care takes the form of elective surgery. For example, uninsured people who are seriously injured in automobile accidents or who have gone into labor are more likely to get inpatient hospital care than those with cataracts or chronic hernias.

There is much discussion of uncompensated hospital services being financed by shifting the costs to paying patients. In fact, the primary subsidies supporting the provision of hospital care to nonpaying patients have long come from local and state governments, from philanthropy, and from medical education. Most hospitals that provide high volumes of care to nonpaying patients have relatively small proportions of patients who pay more than their own way on whom to shift costs (Hadley and Feder, 1985). The most generous and systematic subsidization of uncompensated care takes place in six states—Connecticut, Maine, Maryland, Massachusetts, New Jersey, and New York—where such subsidies are both generated and allocated through a hospital rate-regulation mechanism. In several other states, including Florida, Nevada, and South Carolina, subventions of hospital revenues are also applied to generate revenue for subsidization of hospital-provided uncompensated care. Despite this, the National Association of Public Hospitals (NAPH) reports that 7 of the 90 major public hospitals have closed their doors in the last decade and that 50 percent of all public hospitals are running at a deficit. The NAPH attributes this to three factors: declining public payments (e.g., state and local subsidies have declined 15 percent and Medicare revenues have declined 17 percent); more indigent patients (there was a 9 percent increase in the use of outpatient departments and emergency rooms by uninsured patients between 1983 and 1985; slightly more than half of all such visits were by uninsured or self-pay patients, a quarter of which were written off as bad debt); and the rising cost of treating AIDS (acquired immune deficiency syndrome) (National Association of Public Hospitals, 1987). With regard to the last point, the National

Commission to Prevent Infant Mortality cites the increase in the number of ''boarder babies"—many of whom are infants with AIDS for whom there are no homes (their own or foster homes), who, as a consequence, spend weeks or months in hospitals—as a factor in increased costs to such facilities (Medicine and Health, 1987).

Whatever the sources or magnitude of financing for uncompensated care, they have not grown over the last decade nearly as fast as the number of uninsured people has. It appears that the provision of uncompensated care by hospitals is more sensitive to the supply of the subsidy than to the demand for services. As a result, the amount of uncompensated care provided by hospitals, although it has increased substantially in the past several years, has not increased as fast as the number of uninsured people (Feder et al., 1984). Therefore, uninsured homeless individuals are, in a sense, competing with the growing numbers of other uninsured domiciled people for the relatively scarce resource of subsidized hospital services, especially because the homeless population is concentrated in those areas where the demands on hospitals are greatest.


The other principal pillar of the indigent care system has been the freestanding clinics supported by public or philanthropic funds. These sprang up in many cities contemporaneously with the founding of modern hospitals around the turn of the century and retained a special focus on service to the poor.

The expansion of free-standing clinic services for low-income communities was encouraged by the war on poverty in the 1960s, which supported the development of about 125 neighborhood health centers serving approximately 1.5 million people. In 1976 federal support for such centers was merged with various other forms of categorical support for clinic activities under Section 330 of the Public Health Service Act (Davis and Schoen, 1978). Under the Omnibus Budget Reconciliation Act of 1981 (P.L. 97-35), federal financial support both for community-based health centers and specialized maternal and child health care centers was substantially reduced. A number of localities closed or consolidated facilities. Many health centers, public and private, were forced to reduce their budgets and services. Nonetheless, several hundred such centers survive, in both urban and rural areas, supported by federal and local—and in some instances, state—appropriations, as well as Medicaid and Medicare revenues. The performances of such centers are as diverse and variable as all their other characteristics, but most evaluations have been relatively positive. In most instances they do not appear to have supplanted hospital emergency rooms or clinics as providers of care to the poor, but rather to have reached previously unserved populations, and they remain important to the care of indigents in many communities. Few such centers reached out to the homeless before the initiation of the Robert Wood Johnson Foundation-Pew Memorial Trust Health Care for the Homeless program. However, many now constitute important components of health care services for homeless individuals.

The National Health Service Corps

Over the past decade, many of the facilities and programs that conduct health care for the medically indigent have relied on the National Health Service Corps (NHSC) for the recruitment of professional personnel. The NHSC was established in 1974 both as a response to the problems of inadequate physician supply in underserved areas and a perceived need to subsidize professional education for students who might be attracted to serve in those areas. It is now being phased out in the face of budgetary pressures and a perceived physician surplus. The NHSC has placed physicians, nurse practitioners, dentists, and other health care personnel in neighborhood health centers, free-standing clinics, outpatient departments, and other providers of service to medically underserved communities. In several of the sites visited by the committee where health care services were being provided to homeless people, NHSC practitioners played a central role. The overall performance and cost-effectiveness of the NHSC are matters of some dispute, but the committee is pleased to note that the Stewart B. McKinney Homeless Assistance Act of 1987 (P.L. 100-77) contains specific provisions authorizing NHSC placements in sites providing health care to homeless people.

Categorical Programs

Contributing further to the organizational complexities of providing care to low-income persons are hundreds of specialty outpatient programs that offer such services as family planning, alcohol and drug treatment, mental health care, disease-specific services, or services to specific populations, such as migrant workers or native Americans. While some of these programs were not designed specifically to serve the poor, they do, in effect, provide service to low-income people who are members of those groups to whom the programs are targeted. Many of these programs rely heavily on federal funding. In many instances this source of funding has been cut back in the last decade, with some substitution of state or local funding. The availability of any such services and their relevance to the needs of homeless people vary considerably from community to community, and an accurate generalization is impossible. Such service programs are, however, an important part of the indigent care system in some areas and should not be overlooked in any overview of that system.

Additional Barriers to Access for Homeless People

In their evaluation of the benefit status of homeless people seen in 16 of the Robert Wood Johnson Foundation-Pew Memorial Trust Health Care for the Homeless (HCH) projects, Wright and Weber (1987) noted that a determination of benefit status could not be made for 24 percent of the people seen by workers in these projects. Of the 76 percent for whom benefit status was known, about half were receiving some form of financial benefit and about half were not. They concluded that almost all clients of the HCH projects who were eligible for any benefits from public entitlement programs were already enrolled in them and that those who were not enrolled were not simply because they were not eligible. After examining several possible factors that conceivably could affect whether a homeless person receives any existing form of financial assistance (including Medicaid), they concluded that by far the most significant factor was the leniency of eligibility requirements for each program as determined by each state:

The "entitlement problem" for the homeless is not that they do not participate in programs for which they are eligible, but that in many states they are not eligible for programs of assistance regardless of their rather obvious needs. (Wright and Weber, 1987)

Wright and Weber noted that among the 16 projects included in their study, the proportion of homeless clients receiving any benefits ranged from 22 to 82 percent; the basis for this variation was primarily the eligibility standards of the respective states. However, they also noted that "the benefit levels typical of most social welfare programs are so low as to make only a marginal contribution" to the ultimate goal of "getting the client off the streets or out of the shelters and into some reasonably stable housing and social situation."

Ineligibility is only one barrier to access for the homeless; there are others. The extreme case of such barriers to the homeless has been the requirement that individuals have a permanent address for Medicaid eligibility or for receipt of services from certain providers with defined areas of geographic responsibility. Even with the recent passage of federal legislation prohibiting the requirement of a fixed address as a basis for eligibility for entitlement programs, homeless people are still in no position to receive mail notices of clinic appointments, laboratory or x-ray results, or required reappearances for recertification of Medicaid eligibility or appeals of disability determinations. Absence of an address is not the only documentation problem encountered by homeless people. In the shelters and on the streets, where homeless people spend much of their time, personal identification papers are a valuable commodity sought by thieves and entrepreneurs. The exigencies of a homeless existence rarely afford a means of protecting personal documents from theft, loss, or damage from the elements.

Particular problems that bureaucratic obstacles can create for homeless people were frequently reported to the committee during its site visits. One important example is the scheduling of clinic appointments and ancillary services at times that conflict with the availablity of the only daytime meals homeless people can get or with the time they must begin lining up for shelters to ensure that they have a place to sleep that night. The problems of access put a premium on the ability of homeless people to cope with and manage complicated bureaucratic systems and routines. Those abilities are often limited among homeless people in general and not only among those whose capacity is impaired by mental disorders or substance abuse. For example, the lack of a watch can make the keeping of appointments quite difficult. The use of public transportation, a frequent source of frustration even for the average urban commuter, presents a greater impediment to a person who has money for only one fare and cannot afford to make mistakes in matters of transfers or routes. Moreover, providers experienced in working with homeless people report that they avoid institutions and bureaucratic procedures, often because of prior negative experiences (Brickner et al., 1985). In addition, in many instances service providers are not interested in efforts to reduce these barriers. Homeless people can appear as extremely unattractive potential clients that institutions or professionals may try to avoid. At the most basic level, homeless people often are physically unattractive, unwashed, and lack clean clothes. They may not be compliant, in part because of the exigencies of a homeless existence. Providers often feel that they will not or cannot follow through with therapeutic or self-care regimens (Brickner et al., 1985).

Many homeless people, with or without identifiable mental illness, are passively resistant to service provision, including health services (Vicic and Doherty, 1987). Successfully engaging such persons for purposes of diagnosis and treatment involves extra efforts on the part of health care providers. Many providers may not be sufficiently motivated or knowledgeable to undertake such efforts. The health care system in general, and providers of service to poor people in particular, are most responsive to those patients who are most motivated, demanding, or both. Providers of services tend to react negatively to those who appear to be uncooperative or uncompliant.

Mental Health Care

Approximately one-third of all homeless people show symptoms of mental illness. The barriers to access discussed above generally apply also to this group of homeless people. In most respects, especially with regard to the reluctance of health care institutions and providers to treat the homeless, the barriers are even greater.

There are, however, two major exceptions to this observation. The mentally ill among the homeless are theoretically more likely than those without such a disability to be eligible for SSI benefits (and, therefore, Medicaid). The SSI program specifically provides for benefits to the mentally disabled. This greater access to SSI eligibility is, however, somewhat illusory because the standards by which disability is determined are open to a wide range of interpretations. The extent to which such interpretations can determine eligibility was clearly seen during 1981-1984 when, in response to what was intended to be simply a clarification of such standards, thousands of mentally disabled people were dropped from the SSI rolls. Later restoration of most of them did not completely allay misgivings about the operation.

The potential of the mentally ill for SSI eligibility may be helpful, but the health care benefits received to support treatment for mental illness are generally inadequate, if not inappropriate. In those states in which Medicaid programs fund mental health care (not all do), it is usually only for inpatient care, and then it is only for a very limited period of time (usually 30 days). The prospective patient, therefore, is in a situation of being declared eligible for benefits based on a disability for which treatment is either poorly funded or not funded at all. The concept that such treatment might ultimately enable the individual to move on to productive employment appears to have been sacrificed to cost-containment measures.

Government, especially at the state level, has long accepted a primary responsibility for the provision of mental health services. The creation of large state-operated mental health facilities goes back to the middle 1800s, and many states mandate such a responsibility in their state constitutions. A role for the federal government is more recent, having largely begun with the report of the Joint Commission on Mental Illness and Health, Action for Mental Health, in 1961 and the various pieces of legislation that sought to implement the most significant recommendations of that commission. The general intent of that legislation was to transfer mental health care from the large state-operated hospitals to small community-based mental health centers. This, as much as the discovery and widespread use of psychotropic drugs, was a key element in what is now called deinstitutionalization.

At the time of passage of the Mental Retardation Facilities and Community Mental Health Centers Construction Act of 1963 (P.L. 88-184), it was anticipated that 2,500 community mental health centers would be constructed throughout the United States. At best, only about 700 of these centers ever opened (President's Commission on Mental Health, 1978). Unlike the general health care system, which has a seemingly adequate supply of hospitals, the mental health care system severely lacks the most basic units for treatment—posing the ultimate barrier to access.

Despite substantial increases in state spending on mental health, little has been done to increase the supply of community-based mental health care centers. Of the approximately $6 billion spent annually on state-operated mental health care, more than $4 billion, or approximately 70 percent, is spent on state institutions (Rubenstein, 1986). Although to some extent this allocation of expenditures is the result of the effort to upgrade the quality of care in state institutions that accompanied the effort to transfer care to community-based facilities, the net effect on the mentally ill person in the community remains the same: there is no place to go. In the competition for such a scarce resource as community-based treatment, the homeless person is in a particularly uncompetitive position.

There have been some recent developments toward increasing access to mental health care services for the homeless. One of the most notable programs is the Community Support Program (CSP), which is directed by the National Institute of Mental Health. CSP was developed to provide time-limited demonstration grants in support of efforts to find new approaches for providing services to the homeless mentally ill. The Stewart B. McKinney Homeless Assistance Act of 1987 (P.L. 100-77) authorized an additional $10 million for this program for fiscal year 1987 as well as $30 million for a new block grant program for services to homeless people who are chronically mentally ill.

The mentally ill among the homeless are quite noticeable, and this has raised anew old issues regarding a civil commitment to treatment facilities. The current basis of commitment in most states can be described as a clear and present danger to him- or herself or others. Even in those states that have a broader basis for commitment (e.g., an inability to perform life-sustaining activities), such factors as budget constraints, fear of legal reaction, or a desire to provide treatment in the least restrictive environment have left the clear and present danger criterion as the basis for commitment. As a consequence, a new approach, outpatient commitment, has been proposed (American Psychiatric Association, 1987). Outpatient commitment refers to the procedure by which a court orders treatment in the community for a person who does not meet the standard for civil commitment to an institution (Kanter, 1987). Such an approach is based on an assumption that there are programs in existence that are not only prepared to treat people so committed but that have the facilities and space to do so. That is not the case in most jurisdictions. Especially troublesome are reports of attempts to commit mentally ill people to outpatient services and then require that they pay for those services (Stefan, 1986). The penalties that would be imposed on someone who is committed by the courts to outpatient services and who then refused to pay for such services have not yet been identified.


As many as 40 percent of the homeless adult men in some communities are veterans, and many of those are Vietnam-era veterans. Some may be eligible for services from the largest public health care system in the United States: the Veterans Administration (VA).

The VA health care system includes 172 hospitals, 229 outpatient clinics, 117 nursing homes, 16 domiciliary facilities, and a range of contracted services, directly employing more than 200,000 people at an annual expense of almost $10 billion. In fiscal year 1986, this system provided almost 1.5 million inpatient stays and more than 20 million outpatient visits. It has particularly extensive services for alcoholism and drug dependence, operating 103 specialized alcoholism programs and 51 specialty drug dependence programs (Veterans Administration, 1987).

Veterans with an unblemished discharge status are automatically eligible for free VA services for the treatment of service-connected illnesses or disabilities; free or subsidized services are also available to low-income and elderly veterans. Although they are more stringent than they have been in the past, the income eligibility criteria for VA medical services remain substantially more generous than those of other publicly supported benefits.

Yet, it is widely believed that relatively few homeless people take advantage of the VA health care services to which they are at least theoretically entitled. In part as a response to this perception, the VA has undertaken a number of initiatives to expand services to homeless veterans. Five million dollars has been appropriated for 43 demonstration projects for the community-based treatment of the chronically mentally ill, with particular emphasis on outreach to homeless veterans. Additional funds were appropriated in mid-1987 for the development of new and expanded domiciliary services for homeless veterans, and a number of VA facilities and programs have begun to undertake more aggressive outreach programs to homeless veterans, as mandated by the Omnibus Budget Reconciliation Act of 1986 (P.L. 99-509).

The interim report of that effort, released in October 1987 and covering the first 4 months of the Homeless Chronically Mentally Ill (HCMI) outreach project, partially confirmed the belief that homeless veterans do not—for various reasons—avail themselves of VA medical services (Rosenheck et al., 1987). Although 97 percent of the 6,342 individuals contacted during the first 4 months appeared to be eligible veterans (i.e., veterans who had been honorably discharged), of those for whom intake assessments were completed (N = 4,010), only 18 percent were receiving any financial support from the VA. With regard to access to VA health and mental health care, a sample of 727 veterans contacted through the HCMI program showed that 75 percent of the sample had three or fewer contacts with VA mental health outpatient services during the 6 months immediately prior to contact with the outreach program; 42 percent had no contact with any medical or psychiatric outpatient providers, but 38 percent did have at least one admission to some form of inpatient care.

VA health care facilities are often located at some distance from where homeless people congregate. This presents a serious issue of access, especially if there is no specific outreach program to homeless veterans. This situation is further exacerbated by the VA's system of priority eligibility; who actually receives care is based on the level of disability, whether it is service connected, and the capacity of the specific VA facility in question (U.S. Department of Health and Human Services, 1987). The recent initiatives of the VA are particularly encouraging, but the gap between current and potential services to homeless veterans remains considerable. In addition, the extent of the disability by which one qualifies for full benefits has been raised, making it more difficult for some veterans to qualify. Unless such veterans live in a state where their disability qualifies them for Medicaid, they may be without health care benefits. One especially troublesome issue was reported to the committee during its site visits: Veterans whose disabilities are such that they receive no or only partial VA benefits may be denied non-VA benefits (Medicaid, SSI, or general assistance), on the erroneous assumption, by those responsible for determining eligibility for non-VA programs, that the financial relief for such individuals should come from the VA. The VA HCMI project found that more than half (54.8 percent) of the veterans seen in the sample of intake assessments were receiving no benefits from any source.

One final—and substantial—barrier to access to care for veterans comes in the nature of their discharge from military service. If their military discharge was anything other than honorable (i.e., less than honorable or dishonorable), they are effectively barred from receiving VA benefits.

The VA has a system by which the nature of a discharge can be reviewed and possibly upgraded, but it is cumbersome and requires extensive documentation. It is also not well publicized among veterans, especially disaffiliated homeless veterans. Each of these are specific and serious obstacles to the receipt of health care services by homeless veterans. Outreach and educational programs, such as those that are now being initiated by the VA, as well as counseling programs, which have been conducted for many years by such organizations as Vietnam Veterans of America and Swords to Plowshares, could be highly effective in resolving veterans' access to health care. Ultimately, the success of such efforts depends on the criteria—and the interpretation of those criteria—by which requisites for upgrading of discharge status will be evaluated.


Homeless people, who are predominantly low-income, uninsured residents of mostly low-income communities, share with other low-income residents of such communities a range of difficulties in getting health care when they need it. The most significant barriers to access are financial. At the same time, the homeless encounter a range of additional barriers to health care.

In order to develop service programs to meet the health care needs of homeless people, it is necessary to address both the general issues of access to care and the special problems encountered by homeless people in obtaining access. Financing mechanisms are obviously the first precondition, whether they are through improving eligibility for Medicaid or other benefit programs or through procurement of direct subsidies. Financing, however, while a necessary condition, is not sufficient. Problems of scheduling, transportation, and negotiating systems must also be solved. The practices and attitudes of providers, along with the special problems posed by the passivity, isolation, and resistance of homeless people, many of whom have had negative experiences with health care institutions, must also be taken into consideration. How some of these problems have been met in a number of specific programs is the primary content of the next chapter.


  • Aday, L. A., G. V. Fleming, and R. Andersen. 1984. Access to Medical Care in the U.S.: Who Has It, Who Doesn't. Chicago: Pluribus Press.
  • American Hospital Association. 1986. Hospital Statistics: 1985. Chicago: American Hospital Association.
  • American Psychiatric Association. 1987. Involuntary Commitment to Outpatient Treatment: Report of the Task Force on Involuntary Outpatient Commitment. Task force report no. 26. Washington, D.C.: American Psychiatric Association.
  • Brickner, P. W., editor; , L. K. Scharer, editor; , B. Conanan, editor; , A. Elvy, editor; , and M. Savarese, editor. , eds. 1985. Health Care of Homeless People. New York: Springer-Verlag.
  • Davis, K., and C. Schoen. 1978. Health and the War on Poverty: A Ten-Year Appraisal. Washington, D.C.: Brookings Institution.
  • Davis, K., M. Gold, and D. Makuc. 1981. Access to health care for the poor: Does the gap remain? Annual Review of Public Health 2:159-182. [PubMed: 6756423]
  • Feder, J., J. Hadley, and R. Mullner. 1984. Poor people and poor hospitals: Implications for public policy. Journal of Health Politics, Policy and Law 9(2):237-250. [PubMed: 6491229]
  • Freeman, H. E., R. J. Blendon, L. H. Aiken, S. Sudman, C. F. Mullinix, and C. R. Corey. 1987. Americans report on their access to care. Health Affairs 6(1):6-18. [PubMed: 3583219]
  • Hadley, J., and J. Feder. 1985. Hospital cost shifting and care for the uninsured. Health Affairs 4(3):67-80. [PubMed: 4065809]
  • Holahan, J. 1984. Paying for physician services in state Medicaid programs. Health Care Financing Administration Review 5(3):99-110. [PMC free article: PMC4191359] [PubMed: 10310601]
  • Joint Commission on Mental Illness and Health. 1961. Action for Mental Health: Final Report. New York: Basic Books.
  • Kanter, A. S. 1987. Legal barriers to access: The unmet health care needs of homeless people. Mental Health Law Project, Washington, D.C. Paper prepared for the Institute of Medicine, Washington, D.C.
  • Kovar, M. G. 1987. Letter to the publisher: Questioning the reported decline in use of ambulatory care. Health Affairs 6(Winter):156-157. [PubMed: 3428850]
  • Lee, P., and C. Korenbrot. 1987. The Prevention of Low-Weight Births: A Policy Research Analysis. San Francisco: Institute for Health Policy Studies, University of California, San Francisco.
  • Medicine and Health. 1987. Panel seeks help for AIDS babies. Medicine and Health 41(44):2.
  • National Association of Public Hospitals. 1987. America's Health Safety Net. Washington, D.C.: National Association of Public Hospitals.
  • National Center for Health Statistics. 1988. Health Interview Survey. National Center for Health Statistics, Washington, D.C. (Unpublished.)
  • New York Times. December 1, 1987. Poll finds Reagan support down but Democrats still lacking fire. Al.
  • President's Commission on Mental Health. 1978. P. 319 in Task Panel Reports Submitted to the President's Commission on Mental Health, Volume II, Appendix. Washington, D.C.: U.S. Government Printing Office.
  • President's Commission for the Study of Ethical Problems in Medicine and Biomedical and Behavioral Research. 1983. Securing Access to Health Care: The Ethical Implications of Differences in the Availability of Health Services, Volume l: Report. Washington, D.C.: U.S. Government Printing Office.
  • Robert Wood Johnson Foundation. 1987. Special Report on Access to Medical Care, no. 2. Princeton, N.J.: Robert Wood Johnson Foundation.
  • Robertson, M. J., and M. R. Cousineau. 1986. Health status and access to health services among the urban homeless. American Journal of Public Health 76(5):561-563. [PMC free article: PMC1646594] [PubMed: 3963287]
  • Rogers, D. E., R. J. Blendon, and T. W. Moloney. 1982. Who needs Medicaid? New England Journal of Medicine 307(1):13-18. [PubMed: 7043272]
  • Rosenreck, R., P. Gallup, C. Leda, P. Leaf, R. Milstein, I. Voynick, P. Errera, L. Lehman, G. Koerber, and R. Murphy. 1987. Progress Report on the Veterans Administration Program for Homeless Chronically Mentally Ill Veterans. Washington, D.C.: Veterans Administration.
  • Rubenstein, L. 1986. Access to treatment and rehabilitation for severely mentally ill poor people. Clearinghouse Review 20(4):382-393.
  • Sloan, F. A., and J. Valvona. 1986. Uncovering the high costs of teaching hospitals. Health Affairs 5(3):68-85. [PubMed: 3098661]
  • Stefan, S. 1986. Outpatient Commitment Evaluation of Criteria and Programs. (Unpublished; see also Kanter, A. S. 1987.)
  • Stevens, R., and R. Stevens. 1974. Welfare Medicine in America: A Case Study of Medicaid. New York: The Free Press.
  • Sulvetta, M., and K. Swartz. 1986. The Uninsured and Uncompensated Care: A Chartbook. Washington, D.C.: National Health Policy Forum, George Washington University.
  • U.S. Bureau of the Census. 1985. Disability, Functional Limitation, and Health Insurance Coverage: 1984/85. Series P-70, No. 8, Table 10. Current Population Reports. Washington, D.C.: U.S. Government Printing Office. (Cited in U.S. Congress, House, Committee on Ways and Means, 1987, p. 218.)
  • U.S. Congress, House, Committee on Ways and Means. 1987. Backgroundmaterial and data on programs within the jurisdiction of the Committee on Ways and Means. 100th Cong., 1st sess., March 6, 1987.
  • U.S. Congressional Budget Office. 1986. Tabulations of current population surveys, March 1980-March 1986. Washington, D.C.: U.S. Government Printing Office. (Cited in U.S. Congress, House, Committee on Ways and Means, 1987, p. 219.)
  • U.S. Department of Health and Human Services. 1985. Health Care Financing Program Statistics: Analysis of State Medicaid Program Characteristics, 1984. Baltimore: Office of the Actuary, Health Care Financing Administration.
  • U.S. Department of Health and Human Services. 1987. Health Care Coverage by Age, Sex, Race and Family Income: United States, 1986. Vital and Health Statistics, No. 139, September 18, 1987. Washington, D.C.: National Center for Health Statistics. [PubMed: 10286979]
  • Veterans Administration. 1987. Annual Report: 1986. Washington, D.C.: U.S. Government Printing Office;
  • Vicic, W. J., and P. Doherty, 1987. Homelessness: A Medical Viewpoint. Paper prepared for the Institute of Medicine, Washington, D.C.
  • Wright, J. D., and E. Weber. 1987. Homelessness and Health. New York: McGraw-Hill.
Copyright © 1988 by the National Academy of Sciences.
Bookshelf ID: NBK218231


Related information

  • PMC
    PubMed Central citations
  • PubMed
    Links to PubMed

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...