LITMIR.BIZ

Скачать книгу

when the survey was conducted in 1988–1991 to 4.13% when they survey was conducted in 2004,²⁰ almost as high as Link et al.'s (1994) estimate (4.6%) for a parallel measure over a five‐year period. This, of course, is before the Annual Homelessness Assessment Reports that have shown more recent decreases.

      Studies that were smaller than Link's (Toro, Tompsett, et al., 2007), more specialized (Rosenheck & Fontana, 1994) or more local (Culhane, Dejowski, Ibanez, Needham, & Macchia, 1994) have, like Link, found much higher rates when homelessness is measured over a period of time longer than 1 year. The rates are particularly high for young adults (Morton et al., 2018).

The Importance of Time Frames

The proportion of the population that becomes homeless over an extended period such as 5 years or a lifetime is vastly larger than the proportion homeless on any given night. As national telephone surveys and examination of shelter records for individuals and families have shown, most people who become homeless do not remain homeless forever. As the estimates of numbers of chronically homeless people show, only about a quarter of individuals and 5% of families remain homeless for as long as a year.

      In this, homelessness differs from many other social statuses. The proportion of people in the United States who are immigrants does not vary much over a day or a year or a decade. Some people arrive, others leave or die, there are trends over time, but once an immigrant, always an immigrant. Homelessness is more like unemployment—many people experience it briefly, some have repeated bouts, some despair of finding jobs and leave the labor force entirely.

      For homelessness, as for unemployment, it makes sense to ask how many people are in this status on a given day—what epidemiologists call point prevalence, to understand something about the need for services and to monitor trends. But the one‐night estimate serves to minimize the scope of the problem. To truly understand the vastly larger number of people afflicted, estimates over a longer span of 5 years or a lifetime—what epidemiologists call period prevalence—are also important. Surveys of people living in ordinary housing miss anyone currently homeless, so they are not very good at estimating the effects of recent changes in policies to address homelessness. However, surveys do provide a window on the magnitude of the problem and the resources that will be needed to end it.

      Phrases like “the homeless” suggest to the casual reader that people who are homeless are a species apart—just as the tiger does not change its stripes, “the homeless” will remain so—and maybe there is little anyone can do about it. Understanding that people move into and back out of homelessness provides more points of intervention. What policies and practices can prevent people from becoming homeless? What services can speed their exits and prevent them from returning to that state? The last four chapters of this book address these questions.

Challenges Faced by People Experiencing Homelessness

People who become homeless face many challenges. Among the most severe is high risk of mortality, a risk that is especially marked for younger adults. Excess mortality for younger adults comes from drug overdoses, along with excess suicide, homicide, and infectious disease. Older adults tend to die from the same causes (such as cancer and heart disease) as older adults more generally—but 10–15 years sooner (Baggett et al., 2013; Fazel, Geddes, & Kushel, 2014). Unsurprisingly, people who are unsheltered have higher mortality than people who use shelters, and in a study in Massachusetts that followed the same group of people over 10 years (known as a cohort sample), they had 10 times higher mortality than the overall population (Roncarati et al., 2018). People studied while homeless have high rates of infectious diseases, chronic diseases, psychiatric and substance use disorders, and injuries, some due to victimization (Fazel et al., 2014). Across diagnostic categories, they get sick and are admitted to hospitals at earlier ages (Adams, Rosenheck, Gee, Seibyl, & Kushel, 2007).

      Rates of disability among people who experience homelessness are also extremely high. HUD reports rates of disability from all causes, including physical and cognitive disabilities and those due to substance abuse and mental illness. In 2017, the share of sheltered individuals with disabilities was 49%. This is greater than the share of people with disabilities in the U.S. population with incomes below poverty, 32% (Henry, Bishop, et al., 2018). Levels of disabilities are a bit higher for veterans (59%) and substantially lower for adults in families (22%). Both those figures are still higher than in the general population, or even the population in poverty (16% for families).

The data on disability (based on everyone who used shelter over the course of the year) do not distinguish between disabilities associated with mental illness, substance abuse, developmental disability, cognitive impairment, and physical causes.²¹ Nor does HUD look carefully at the way disabilities are identified by the program staff that report data into the information system that is the basis for the HUD estimates. But we think the basic point is accurate: about half (51%) of adults who experience sheltered homelessness over the course of a year as individuals do not have any disability, and more than three quarters (78%) of those who are part of a family with children do not have a disability.

      Studies of mental health and substance abuse problems among people experiencing homelessness find wildly different rates. For example, a systematic review of mental health diagnoses among homeless individuals (excluding families) in wealthy countries, primarily in Europe, found prevalence rates for psychotic illness ranging from 2.8 to 42.3% and of major depression ranging from 0.0 to 40.9%. The best estimates from this review, considering factors such as the size of samples, are that among unaccompanied adults experiencing homelessness, 12.7% currently have a psychotic illness, 11.4% major depression, 23.1% a personality disorder, 37.9% alcohol dependence, and 24.4% drug dependence. Slightly higher rates of alcohol dependence and slightly lower rates of major depression are found in mainland Europe compared to other wealthy countries (Fazel, Khosla, Doll, & Geddes, 2008).

Almost all of the studies were cross‐sectional—that is, they sampled a group of people who are homeless at a particular point in time, rather than a sample of all people who become homeless over a period of time. We have already shown that studies describing people at a particular point in time provide a different picture from studies showing who becomes homeless over a longer time period. (For example, the proportion of families who are homeless on a given night is larger than the proportion homeless over the course of a year because families stay longer, on average.) Why might that matter for understanding rates of disability among people experiencing homelessness?

      Compared to all people who ever become homeless, those who are counted in a point‐in‐time survey include a higher proportion of people who remain homeless for a long time or return repeatedly to that state, that is, of people with chronic patterns of homelessness. People who became homeless a month or 6 months before the survey but who returned to housing quickly are not included; others who became homeless at the same time but remained so are counted. To the extent that problems such as mental illness or substance abuse make it difficult to extricate oneself from homelessness, estimates of those problems will be magnified in cross‐sectional studies (Phelan & Link, 1999). A PIT count is an example of a cross‐sectional study. It minimizes estimates of the number of people who experience homelessness but maximizes estimates of their problems.

      To illustrate, we return to the typologies of homelessness for individuals and families and focus on Philadelphia, where researchers matched records of shelter use to records of treatment for substance abuse (Culhane et al., 2007; Kuhn & Culhane, 1998). For individuals, fewer than 10% of people who entered shelter were long stayers, but they used just over half of the shelter days. That means that, on any given day, one would be likely to find that just over half of the shelter residents were long‐term users, and the long stayers and episodic users of shelters had more problems such as substance abuse than the transitional (short‐stay) shelter users. Among all individuals who entered shelter over 28 months, 37% reported substance problems, and 29% had received substance abuse treatment from a publicly funded source. If a cross‐sectional

Скачать книгу