Homelessness Research Paper

1. Defining Homelessness

Homelessness did not exist as a public policy issue from the end of the Great Depression until the early 1980s. Certainly there were people without homes, or with very inadequate or dangerous homes. In some poor areas of many cities known as ‘skid rows,’ shelters, or ‘missions’ run by religious groups offered a bed and sometimes a meal in exchange for the chance to save a soul. These skid row areas also contained hotels where one could rent a small room, with perhaps a hotplate and a small sink, for a day, a week, or a month depending on how much money one had. In addition, rooming or boarding houses offered inexpensive accommodations.

Before 1980, the US Census classified people as ‘homeless’ if they lived by themselves and not with family members. Thus, all of the single people living alone in skid row areas (and elsewhere) might have been called homeless—a definition quite different from the one in use today, when very large numbers of the general population live alone by choice. Because of definitional problems and the absence of a concept close to what today would be called homelessness, there are no good estimates of the size of the literally homeless population for the 1960s and 1970s. The best guess is that in the late 1970s formal shelters (i.e., missions) for the homeless might have been able to accommodate about 50,000 people at any one time in the entire country, and that the number of literally homeless people probably was no more than 100,000 and could have been considerably less.

Since 1987, official US policy defines people as literally homeless if they sleep in a shelter for homeless people or in places not meant for human habitation such as vehicles, abandoned building, outdoor locations, or transportation facilities or stations. However, the people included in any actual study of the homeless, and therefore the descriptions of homeless populations and estimates of their size, will vary greatly depending on study methods (Taeuber 1991). If studies go only to shelters they will miss all homeless people who do not stay in them. Even studies that include searches of nonshelter locations have a very hard time finding most unsheltered homeless people. The best methods developed to date use services such as soup kitchens and outreach programs to find the nonsheltered homeless, and succeed better than street searches. In some European countries with a right to housing, homelessness is sometimes defined quite differently, as people who come to a housing authority needing housing (which they usually receive). Obviously both definitions of homelessness and methods used to find homeless people will affect estimates of population size, while public policies will affect the types of remedy available and the probable length of homeless careers.

2. Homeless Numbers And Rates

The best point prevalence estimate of the size of the homeless population in the US before the early 1980s was 100,000 or fewer. By 1984 the estimate was up to 250,000 to 300,000, and by 1987 to 500,000 to 600,000. Estimates for 1996 of homeless people using homeless assistance programs (not all homeless people) range from 444,000 for October to 842,000 for February (Burt et al. 2001). All of these rates include any children accompanying a homeless adult, but not runaway and homeless youth.

The best way to compare estimates over time and place is to convert them to rates, which for homeless estimates is usually a rate per 10,000 population. Homelessness rates based on the 1987 estimate of 500,000 to 600,000 homeless people were 19.4/10,000 to 24.3/10,000. Projecting the February 1996 estimate of homeless service users to the whole homeless population assuming that service users during this winter season were 80 percent of the population produces an estimated rate of 38.0/10,000. This significant increase in the estimated rate of homelessness at any given time owes a good deal to the more than 200 percent expansion of homeless shelter capacity between 1987 and 1996 (see below), all of whose residents are classified as homeless.

Annual prevalence rates are a far better vehicle that point prevalence rates for understanding the proportion of a population affected by homelessness over time, because they account for the very large number of homeless spells lasting less than one year. Estimates for New York and Philadelphia indicate that 1 percent (100/10,000) of the entire population of those cities used a homeless shelter during a year’s period (Culhane and Hornburg 1997, pp. 101–34). The same level of homelessness was estimated for the entire US for 1996 by Burt et al. (2001). When considered as a proportion of the poor population, the rate is close to 10 percent experiencing homelessness in a year’s time.

3. Major Characteristics Of Homeless People In The United States

3.1 Demographic Characteristics

In 1996, information from NSHAPC indicated that about one-third of homeless services users were members of homeless families; 23 percent were the children in these families, and 11 percent were their parents. These families comprised 15 percent of all homeless households; 86 percent of which were headed by women. Among family households, 38 percent were white non-Hispanic, 43 percent were black non-Hispanic, 15 percent were Hispanic, and the remainder were from other races ethnicities. Forty-one percent had never married, and only 23 percent were currently married. Educational attainment was very low, with 53 percent having less than a high school education. These families had a mean monthly income of $475, or less than half of the US federal poverty level for a family of three (Burt et al. 1999).

Nonfamily homeless households were 61 percent men by themselves, 15 percent women by themselves, and 10 percent men or women with at least one other adult (Burt et al. 2001). Among nonfamily homeless service users, 41 percent were white non-Hispanic, 40 percent were black non-Hispanic, 10 percent were Hispanic, and 9 percent were from other races ethnicities. Half of these homeless adults had never married, and only 7 percent were currently married. Equal proportions (36–37 percent) had completed high school and had less than a high school education. Mean monthly income was $348 or 51 percent of the federal poverty level for a single person (Burt et al. 1999).

3.2 Other Characteristics

In 1996, less than 10 percent of homeless adults worked in a regular job while homeless. However, 44 percent did some paid work for a month. Homeless adults were almost disconnected from the financial support of family and friends, with only 21 percent receiving money from these sources in an average month. Homeless families were somewhat better connected to public benefits, with 52 percent receiving the primary public welfare benefit of Aid to Families with Dependent Children and associated food and healthcare benefits. About 1 in 10 homeless adults received Supplemental Security Income, a program for people with serious long-term disabilities preventing them from working. Many studies find high levels of alcohol, drug, and mental health problems among homeless populations; the NSHAPC included. During the year before the study, 46, 38, and 45 percent of homeless adults reported alcohol, drug, and mental health problems, respectively. Only one in four did not report any of these problems. Homeless women were more likely to report mental health problems without either alcohol or drug problems, while homeless men were most likely to report having problems only with alcohol and not with drugs or mental health issues.

3.3 Patterns Of Homelessness

Peoples’ experiences of homelessness vary considerably. Many researchers have differentiated three patterns of homelessness—first-time crisis transitional, episodic, and chronic—although the exact classification criteria vary across studies. The first group includes individuals or families who are homeless only once or twice and usually for a relatively short period of time (e.g., less than 3, 6, or 12 months). Often they have been precariously housed and find themselves homeless after an unexpected crisis such as unemployment, divorce separation, or eviction. They usually secure some type of housing relatively quickly and often do not become homeless again. Episodically homeless people are defined by their pattern of cycling in and out of homelessness over significant periods of time. Many spend time in various institutions such as inpatient treatment centers, detoxification programs, and jail. Chronically homeless people have relied on shelters or lived on the streets for many months or years, and usually have multiple barriers to securing stable employment and housing. They may have only a few distinct spells of homelessness, but each one lasts for a very long time.

Point-in-time studies, which overestimate the proportion of homeless people with long spells, usually find that about one-quarter of homeless adults are in the first group, and that half or more have extensive histories of either episodic or continuous homelessness. Studies that are able to observe flows into and out of the homeless condition over time give a different picture, showing only 10 to 20 percent with long-term homelessness and up to half with quite short-term or crisis patterns (Culhane and Hornburg 1997, pp. 101–34).

4. Antecedents Of Or Factors Associated With Homelessness

Both societal and personal factors affect homelessness. The societal factors predict when and where homeless populations will increase or decrease; the personal factors help to predict who will be vulnerable to these societal pressures. The two sides of the equation determining the ability to afford housing are household income and housing costs. In the US, housing costs for the lowest-cost units has increased significantly in the past 2 decades. Also, homelessness most recently became a public issue in the early 1980s, during a time of record-high unemployment (the recession of 1981– 1982) that reduced household income among both low-wage workers and workers in manufacturing, many of whom lost their jobs. The record prosperity of the US during most of the 1990s has been accompanied by the development of a two-tiered labor market that has relegated many people without a college or technical education to low-paying service jobs. At the same time, wealthier households have pushed housing costs to record-high levels. The differential between earning power and housing costs sets the societal stage for homelessness. A public safety net of housing and other programs could assure that households do not founder under such circumstances. Some European countries have such a safety net, but the US does not. Some level of homelessness is the result.

That being said, certain people are more likely than others to become homeless. Poverty is the sine qua non of homelessness. The explanatory value of demographic characteristics such as race ethnicity diminish or disappear in many analyses that simultaneously include income variables plus characteristics such as physical health, childhood experiences, and variables relating to alcohol and drug use and mental health problems. Regression results from an increasing number of studies on both families and singles confirm that sexual or physical abuse by a household member, foster care, or other out-of-home placement in childhood differentiate both homeless families and homeless singles from their housed counterparts. Physical and mental health conditions in adulthood, and early and current alcohol or drug use are also usually found to increase the probability that a person will experience homelessness when other things are held constant.

5. Growth In Services And Evidence Of Effectiveness

5.1 Shelter Housing Capacity

NSHAPC offers the first opportunity in many years to update knowledge about homeless assistance programs from the national perspective. Estimates based on NSHAPC data indicate that the nation’s shelter housing capacity within the homeless assistance system grew by 220 percent between 1988 and 1996, from the 275,000 beds estimated by the US Department of Housing and Urban Development (HUD) for 1988 to more than 600,000 beds in 1996. This growth came on top of the 275 percent increase in shelter capacity between 1984 and 1988 from 100,000 to 275,000 beds. One of the most striking things about the growth of shelter housing capacity is the extent to which the system has gone beyond emergency shelter to include large numbers of transitional housing units, and permanent housing units for disabled people who were once homeless. In fact, relatively little growth occurred in emergency shelter programs, or beds specifically, between 1988 and 1996. HUD estimated 5,400 shelter programs in 1988, with a capacity of 275,000 beds. Virtually all of these programs and their beds supplied emergency accommodation, either through emergency shelters or through vouchering arrangements with hotels and motels. In 1996, NSHAPC data indicate that the number of beds available through more than 8,800 emergency shelter plus voucher distribution programs had grown to only 333,500, or a mere 21 percent over the level in 1988.

In contrast, transitional and permanent housing units exploded, going from basically none in 1988 to about 275,000 in 6,300 programs by 1996. Thus, this entirely new capacity to help people with housing needs beyond emergency shelter was essentially equal, in 1996, to the entire homeless assistance system in 1988. This growth coincided with the initial availability and eventual expansion of federal funding for new types of shelter housing stimulated by homeless legislation.

5.2 Soup Kitchen Meal Capacity

No national data exist to describe soup kitchen and other prepared meal distribution programs for 1987 that are equivalent to HUD’s 1988 survey of shelter capacity. However, a 1987 Urban Institute study (Burt and Cohen 1989) did provide a roughly equivalent look at soup kitchen and mobile food program capacity in cities over 100,000, which can be contrasted to NSHAPC’s estimates for central cities in 1996. Bigcity prepared meal capacity increased almost four-fold between 1987 and 1996, from 97,000 up to 382,100 meals a day. Nationally, NSHAPC estimates are that soup kitchens and mobile food programs expected to serve almost 570,000 meals a day in February 1996, of which one-third were served in suburban urban fringe areas or rural areas outside of central cities.

Unlike most shelter housing programs, users of which are presumed to be literally homeless, soup kitchens and mobile food programs commonly serve many nonhomeless people (although some may have had earlier episodes of homelessness). In addition, people may eat more than one meal a day at these programs. For both these reasons, the estimate of meal service levels in 1996 cannot be interpreted as a level of homelessness. However, it certainly can be interpreted as a level of need. Also, soup kitchen prepared meal programs have not had even one-tenth of the federal financial incentive to increase their service level that has been available to open or expand shelter programs over the past decade. This makes the service growth in the emergency food arena between 1987 and 1996 even more remarkable than the growth of the shelter housing segment of homeless assistance. It attests to the responsiveness of many communities’ emergency capacity, as well as to the ongoing problem of hunger even in a decade when most Americans enjoyed remarkable prosperity.

5.3 Evidence Of Effectiveness

Effectiveness should only be assessed against expectations that are realistic to the programs and services examined. If the goal is to supply shelter and food, then shelters and soup kitchens are effective in meeting that goal. If the goal is to end, or prevent, homelessness, then the growth in shelter capacity in the US has probably had the opposite effect. This perverse result arises from the definition of homelessness, which includes residence in a homeless shelter. It is unclear whether the underlying level of need—the pool of people desperate enough to seek shelter from the homeless service system—grew in the 1990s. But the expanded shelter capacity gave such people a place to go that did not exist in the 1980s, and they have used it.

In 1999, a federally sponsored symposium featured papers summarizing ten years of research on the effectiveness of services for homeless people (Fosburg and Dennis 1999). Most of the findings revealed that intermediate program goals were being achieved. For instance, outreach programs were able to engage homeless people with serious disabilities, and eventually connect them to services. Case management programs were also able to connect homeless people to services. Psychiatric services were able to help mentally ill homeless people with their mental health problems. However, on average, none of these programs were able to end homelessness for their participants unless they supplied housing (Shinn and Baumohl 1999). If they supplied housing, especially with supportive services as required, programs could help even the most disabled homeless people leave the streets and maintain themselves in stable residential situations.

6. Future Directions For Policy And Research

In the US the argument has come full circle after almost 2 decades. The notion that homelessness was a problem of ‘housing, housing, and housing’ was made forcefully in the early 1980s, and has now proved to be quite accurate. Service programs have helped many homeless people in many ways. But to help people leave homelessness for good, or to divert people from entering into homelessness, the resources required are housing resources. These include help with landlords, back rent, evictions, and move-in expenses as diversion outlays, and actual housing as a means to end homelessness. A big question exists as to the value of supportive services to accompany the housing, once the housing itself is supplied. Research on this issue is summarized by Shinn and Baumohl (1999). Some research, mostly on homeless families, suggests that housing is the key and that services add little or nothing. Other research, more focused on chronically homeless people with significant alcohol, drug, and/or mental health problems, indicates an important role for supportive or supervisory services to accompany housing.

Future research could profitably take two directions. The first direction would focus on what supports or services made available just before a homeless episode or at the moment of its precipitating event could avert the transition to literal homelessness by stabilizing the situation or simply redefining the issue. For instance, some community support services for persons with chronic mental illness offer ‘respite’ facilities where a person can stay for several days during a crisis in their illness. Respite stays avert complete dissolution of the person’s life in the community, including loss of housing. It would be important to learn how these types of interventions can work best, for whom, and when. The second direction for research would focus on what services in addition to housing would make the most difference for people who are already homeless. Some communities in the US have already shifted resources from emergency shelter to affordable and supported housing, actually reducing the need for emergency shelter beds. More attention needs to be focused on developing a range of such approaches, because it should be clear that continuing to increase the capacity of the homeless service network will only serve to bring more and more people into the category of homeless.

Bibliography:

1. Burt M R, Aron L Y, Douglas T, Valente J, Lee E, Iwen B 1999 Homelessness—Programs and the People They Serve: Summary and Technical Reports. Interagency Council on the Homeless Department of Housing and Urban Development, Washington, DC
2. Burt M R, Aron L Y, Lee E 2001 Helping America’s Homeless: Emergency Shelter or Affordable Housing? Urban Institute Press, Washington, DC
3. Burt M R, Cohen B E 1989 America’s Homeless: Numbers, Characteristics, and Programs that Serve Them. Urban Institute Press, Washington, DC
4. Culhane D P, Hornburg S P (eds.) 1997 Understanding Homelessness: New Policy and Research Perspectives. Fannie Mae Foundation, Washington, DC
5. Fosburg L B, Dennis D L (eds.) 1999 Practical Lessons: The 1998 National Symposium on Homeless Research. US Department of Housing and Urban Development and US Department of Health and Human Services, Washington, DC
6. Shinn M, Baumohl J 1999 Rethinking the prevention of homelessness. In: Fosburg L B, Dennis D L (eds.) Practical Lessons: The 1998 National Symposium on Homeless Research. US Department of Housing and Urban Development and US Department of Health and Human Services, Washington, DC
7. Taeuber C M (ed.) 1991 Enumerating Homeless Persons: Methods and Data Needs. US Department of Commerce, Bureau of the Census, Washington, DC