Counting and categorizing the homeless is a difficult task, both practically and theoretically. Every year or so, the United States Department of Housing and Urban Development (HUD) prepares a report for the Congress—called the Annual Homeless Assessment Report (AHAR)—on the state of homelessness in the country. It attempts to do just that: demograph the homeless population across the entirety of the USA. HUD also publishes a more frequent quarterly report for its Homeless Pulse Project to carefully monitor the trends in homelessness in cooperating selected communities, and also trends which might affect social policy and funding. HUD funds facilities for the homeless, poorer families, and others, so it is rightly concerned about these numbers.
Counting and categorizing the homeless population is a very delicate statistical measurement. Who exactly is homeless? There does exist a working, legal definition of a homeless person, so it seems fairly easily calculable. But it’s controversial. Two basic methods are employed for data collection: Point in Time (PIT) counting, and data collected via a computerized Homeless Management Information System (HMIS).
In a PIT count, one selects a particular night and goes out in the streets to count and demograph those who are without permanent abode. It’s totally hands on. Many cities in the USA do their own PITs too, separate from the HUD assessments. Obviously, numbers yielded are just estimates. Alternatively, the HMIS records people who are staying in homeless shelters, transitional housing programs, detoxification programs, holding programs, post-incarceration re-entry programs, and in other places. In an HMIS there is a government standard to determine which data are kept in the system. Controversies surrounding the accuracy of both measurement tools abound, but they seem to be the best that anyone has come up with for the time being.
The last AHAR to the Congress was published in July 2009, and reflects statistics based on data from 2008. It’s entitled “The 2008 Annual Assessment Report to Congress”. Some key findings are as follows:
∑ On a single night in January 2008, there were 664,414 sheltered and unsheltered homeless persons nationwide.
∑ Nearly 6 in 10 people who were homeless at a single point-in-time were in emergency shelters or transitional housing programs, while 42% were unsheltered on the “street” or in other places not meant for human habitation.
∑ About three-fifths of the people homeless on a single night were homeless as individuals (62%), while two-fifths (38%) were homeless as part of a family. Family members were much less likely than individuals to be unsheltered. About 27% of all homeless family members were unsheltered on the night of the point-in-time count, while almost half of homeless individuals were unsheltered.
∑ One-day PIT counts of homelessness changed little between 2007 and 2008: the total number of homeless persons decreased by about 1 percent or 7,500 people.
∑ The January 2008 PIT estimate of chronic homelessness across the United States is 124,135 persons, or 30% of all homeless individuals. The PIT count of chronically homeless persons in 2008 is nearly identical to the count in 2007.
∑ The concentration of homeless persons in a state—or the estimated number of homeless persons as a percent of the state’s total population—varies considerably across the United States.
∑ Homelessness is heavily concentrated in urban areas, and especially in certain cities. For instance, in both 2007 and 2008, one in five people homeless on a single night in January were in Los Angeles, New York, or Detroit.
∑ About 1.6 million persons used an emergency shelter or a transitional housing program during the 12-month period from October 1, 2007 through September 30, 2008, suggesting that 1 in every 190 persons in the United States used the shelter system.
∑ The nation’s sheltered homeless population over a year’s time included approximately 1,092,600 individuals (68%) and 516,700 persons in families (32%).
∑ The most common demographic features of all sheltered homeless people are: male, members of minority groups, older than age 31, and alone.
∑ More than two-fifths of sheltered homeless people have a disability.
∑ About two-fifths of people entering an emergency shelter or transitional housing program during 2008 came from another homeless situation (sheltered or unsheltered); two-fifths came from a housed situation (in their own or someone else’s home); the remaining one fifth were split between institutional settings or other situations such as hotels or motels.
∑ Most people had relatively short lengths of stay in emergency shelters: three-fifths stayed less than a month, and a third stayed a week or less. Stays in transitional housing were longer: about 40 percent stayed 6 months or more in 2008. Nearly 1 in 5 people in transitional housing was there for the entire year.
The entire 2009 AHAR report can be found online at HUD’s website (hudhre.info), or by following this link: http://www.hudhre.info/documents/4thHomelessAssessmentReport.pdf
An examination of a more recent HUD report, “The Homelessness Pulse Project: Fourth Quarterly Report”, published in May 2010, offers what is meant to be a more up-to-date look at specially selected communities (e.g., Phoenix, New York City, Washington D.C., et al.) as indicators of the state of the nation. The Pulse Project also endeavors to scrutinize the effects of economic crises on homeless patterns. This report utilizes and compares data from the end of December 2009—the final quarter of that year—with information from previous quarters in 2009.
Some key findings, from this more recent research:
∑ The total end-of-December PIT count marks a slight decrease in the total sheltered homeless population from the previous quarterly counts.
∑ The total count of sheltered homeless individuals in both Q3 and Q4 decreased by 1 percent from Q3, following a 4% increase the previous quarter.
∑ There were 29,694 newly sheltered homeless persons (“new clients”) reported for October through December, an 8% decrease from the 32,437 reported for the quarter that ended in September.
∑ Five of the eight sites reported substantially more new individuals than new persons in families among newly sheltered homeless persons.
∑ Overall, 32% of new clients in Q4 were children (including 1% unaccompanied youth).
∑ This is the first Pulse report to collect information on prior living arrangements for new clients. The data—which was compiled only from study in New York City—show that the clear majority of newly sheltered homeless persons came directly from “housing situations” (55%), including owned or rented units, permanent supportive housing, and staying with family or friends. Of this group, most new clients previously stayed with family or friends (about two-thirds of new clients coming from housing situations), while a significant proportion previously came from a rented unit (27% of new clients coming from housing situations). Only 2% of the new clients sought shelter or transitional housing after spending the night in a unit that they owned.
This entire 2010 Pulse report can also be found online at HUD’s website (hudhre.info), or by following: http://www.hudhre.info/documents/HomelessnessPulseProjectMay10.pdf
Some salient points can be gleaned from these reports. Most homeless people are single males without families. Many have mental illnesses, disabilities, or substance abuse issues. There is an increasing number of families and children going homeless for the first time. Furthermore, many individuals are going homeless for the first time after living with relatives as a guest.
While homeless individuals account for less than 1% of the population in each state, the number of people who are homeless as a proportion of the overall population varies, as homelessness is primarily concentrated in major urban centers. The recent HUD and Pulse reports emphasize that while people from a similar demographic as in the past still comprise the ranks of the chronically homeless, the number of ”first time” homeless has risen in recent years. This includes families and children, as well as individuals such as military veterans and recent victims of domestic violence.
HUD’s data collection techniques represent the best—or at least the most comprehensive—tool we have, besides generous human hearts, to try to get a handle on and ameliorate the American homeless conundrum. One is left with hope, but also recalls the poignant words of the 19th century American writer Mark Twain: “The are three kinds of lies: lies, damned lies, and statistics”. Yet despite any inaccuracies in the HUD reports, the stark truth of our society’s need to solve homelessness remains.