HUD Presents 2nd National Report to Congress on Homelessness in US.

Following is the Executive Summary of the report with footnotes omitted and reformatting for copying purposes. The report can be viewed in it's entirety at: http://www.hudhre.info/documents/2ndHomelessAssessmentReport.pdf

The U.S. Department of Housing and Urban Development (HUD) is pleased to present this second national report to Congress on homelessness in America. The first Annual Homeless Assessment Report (AHAR) was submitted in February 2007. These reports were developed in response to a series of Congressional directives beginning with the FY 2001 HUD Appropriations Act. In that year, Congress directed the Department to assist communities to implement local Homeless Management Information Systems (HMIS) and required every jurisdiction to have client-level reporting within three years. Senate Report 106-410 noted that HMIS data could be used to develop an unduplicated count of homeless people and to analyze the use and effectiveness of homeless assistance services. To that end, Congress further charged the Department with collecting and analyzing HMIS data from a representative sample of communities in order to understand the nature and extent of homelessness nationally.

The second AHAR makes use of two primary data sources. The first source is HMIS data on the number, characteristics, and patterns of shelter use among sheltered homeless persons— or persons who used emergency and transitional housing—during a six-month period from January 1 through June 30, 2006. The data were obtained from a nationally representative sample of communities. A total of 58 sample sites participated in the second AHAR, including 49 communities that participated in the first AHAR and 9 new sample communities that were not able to provide data for the first report. Because some sample communities are still working to secure the participation of homeless assistance providers in HMIS, not all could provide data for this analysis (or could provide only partial data). As a result, the estimates provided in this report have large confidence intervals (i.e., sampling error).

In addition to the sample communities, 16 communities, or “contributing communities,” that were not part of the original sample met the minimum requirements for participation and volunteered to provide their data for this second report. These communities, or “contributing”
sites, have advanced HMIS systems, and several had participated in the first AHAR. Their data help to improve the reliability of the national estimates.

The report also makes use of data provided by all Continuums of Care (CoCs) as part of their 2006 HUD application for funding. The CoC application data contain information on sheltered and unsheltered homeless persons on a single night in January 2006. While only for a single night, these point-in-time (PIT) data complement the HMIS data because they provide information on the number of unsheltered homeless persons and on the national inventory of homeless shelter beds.

The remainder of this Executive Summary reviews the key topics addressed in the AHAR:

l The number of homeless persons based on point-in-time counts;
l The number & characteristics of sheltered homeless persons based on longitudinal HMIS data;
l The nation’s capacity to house homeless persons;
l Where homeless persons receive shelter; and
l The patterns of shelter use in emergency shelter and transitional housing.

The Number of Homeless Persons at a Point in Time - According to CoC application data, the total number of homeless persons reported on a single night in January 2006 was 759,101. At this point in time, more than half of the nation’s homeless population (56 percent or nearly 428,000 persons) were sheltered, while 44 percent (331,000 persons) were unsheltered. Overall, these numbers represent a slight decrease when compared to the PIT data reported by the CoCs in 2005 (from 763,010 in 2005 to 759,101 in 2006)

This change is comprised of a decrease (-13,700) in the total number of unsheltered homeless persons, offset somewhat by an increase (+ 9,800) in the number of sheltered homeless persons. Of the nearly 428,000 people in shelter, approximately 52 percent were persons in households without children, while approximately 48 percent of the sheltered homeless were persons in households with children. By contrast, unsheltered homeless persons were more than twice as likely to be in households without children. Nearly 70 percent of unsheltered persons were in households without children, while approximately 30 percent were persons in households with children. Compared to data reported in the first AHAR, we find that the distribution of sheltered and unsheltered persons by household type remains essentially unchanged.

Point-in-time data from CoC applications also provide information about sheltered homeless subpopulations, including the number of persons who are chronically homeless. Ending chronic homelessness has been a goal of the Administration for several years. A chronically homeless person is defined as an unaccompanied homeless individual with a disabling condition who has either been continuously homeless for a year or more or has had at least four episodes of homelessness in the past three years. To be considered chronically homeless, a person must have been on the streets or in emergency shelter (i.e., not in transitional or permanent housing) during these stays.

Based on their PIT counts, CoCs reported a total of 155,623 chronically homeless people in their jurisdictions in January 2006. This represents approximately 21 percent of the total sheltered and unsheltered homeless population. Approximately 66 percent of chronically homeless individuals in January 2006 were unsheltered homeless persons, and 34 percent were sheltered homeless persons. Compared to 2005 data reported in the first AHAR the number of chronically homeless persons declined by 11.5 percent (from 175,914). The decline could be partially attributed to HUD’s ongoing efforts to address the special needs of this subpopulation by developing permanent supportive housing and providing local communities with technical assistance guidance for developing effective interventions. It may also be attributed to improved techniques among local communities to capture this information accurately.

The HMIS data provided by the AHAR sample allow for estimation of the number and characteristics of people using homeless services over time. The population of people using homeless services over time is different from the population at a single point in time. Point-in-time estimates capture a higher share of homeless individuals and families who use shelters or transitional housing for long periods of time and underrepresent people whose homelessness is episodic (cycling in and out of shelters) and people who have single, brief episodes of homelessness. Thus, HMIS data can provide a more accurate picture than point-in-time estimates of the characteristics and shelter use patterns of people who experience homelessness over a period of time.

Based on the HMIS data provided by the national AHAR sample, more than 1,150,000 total persons used emergency shelter and/or transitional housing nationwide from January through June 2006. Nearly three-quarters (73 percent) of all shelter users during the six-month time period are homeless as individuals or in households without children. About one-quarter (27 percent) are members of households with children. By comparison, a much larger proportion of the nation’s poor (65 percent) and the total U.S. population (55 percent) are persons in households with children.

Other key findings about sheltered homeless persons based on six months of HMIS data include:

The majority of all shelter users (53 percent) are single adult males. By comparison, single adult men constitute just 23 percent of the U.S. population and 16 percent of the poverty population.


Children represent roughly 20 percent of all people who use the shelter system. This includes unaccompanied youth and children in households with adults. Although this is a lower percentage than that of children among the U.S. poverty population (35 percent), the number of children who not only are poor but also become homeless is a cause for concern.


Homelessness disproportionately affects minorities, especially African Americans. Minorities constitute one-third of the total U.S. population and about half of the poverty population, but about two-thirds of the sheltered homeless population. African-Americans are heavily overrepresented in the sheltered homeless population, representing about 44 percent of the sheltered homeless population but 23 percent of the poverty population and only 12 percent of the general population.


Fourteen percent of all homeless adults who accessed a shelter during the six-month time period are veterans. While underrepresented among the poverty population, veterans are overrepresented in the homeless shelter population when compared to the general population.


A significant proportion of the sheltered homeless population is disabled. Sheltered homeless adults are more than twice as likely to have a disability when compared to the general U.S. population. Approximately 38 percent of adults who used a shelter between January 1 and June 30, 2006 had a disabling condition compared to 30 percent of the poverty population and 17 percent of the total U.S. population.

The Nation’s Capacity to House Homeless Persons

According to 2006 CoC application data, there are approximately 406,586 emergency and transitional year-round beds nationwide. About one-half of the total year-round housing inventory (206,877 beds or 51 percent) is in emergency shelters, and the remaining inventory (199,709 or 49 percent) is in transitional housing programs. The mix of available year-round beds differs slightly across household types. There are more family beds in transitional housing (about 103,743 beds) than in emergency shelters (95,301 beds), and conversely, there are more individual beds in emergency shelters (111,576 beds) than in transitional housing (95,966 beds).
The 2006 inventory also includes approximately 21,769 seasonal beds and 55,047 overflow/voucher beds, which are used sporadically throughout the year depending on weather conditions and demand. If these beds are added to the total number of year-round shelter beds in emergency and transitional housing programs, the nation’s peak bed capacity for homeless persons is about 483,402 beds.

In addition to funding emergency shelter and transitional housing beds, HUD continues to encourage communities to develop permanent supportive housing for formerly homeless persons. Overall, there are about 196,626 permanent supportive housing beds in the nation’s bed inventory. Approximately 56 percent of the beds (109,351) are in projects serving unaccompanied individuals, while the rest (87,275) are in projects serving families.

Where Homeless Persons Receive Shelter - Homelessness is, in general, concentrated in central cities. Based on data from the 2006 AHAR sample, approximately 75 percent of homeless persons are in central cities rather than in suburban or rural areas. This is roughly double the proportion of the poverty population in central cities and three times the proportion of the U.S. population in central cities.

Mobility patterns among homeless people most likely account for much of these differences. A 1996 study of people using homeless assistance services indicates that only 28 percent of homeless persons began their homeless spell in a central city (which is only a little higher than the share of the population living in central cities), and 44 percent of homeless persons left the community where their current homeless spell began. The AHAR data suggest that much of this mobility among homeless persons consists of moving from suburban or rural areas to central cities. There are many possible reasons for such mobility, including movement to more densely populated areas to find jobs, to be closer to relatives who may provide support, or to access the greater variety of homeless residential and supportive services that may be available in a larger city.

Patterns of Shelter Use in Emergency Shelter and Transitional Housing - More than three-quarters of all those served by homeless residential programs between January and June 2006 used emergency shelters only. About 18 percent used transitional housing programs only, and a small share (3.5 percent) accessed both types of residential services.

The length of stay in emergency shelters and transitional housing is very different for families with children than for individuals. For emergency housing only, about 19 percent of the persons in households with children stayed a week or less compared to 40 percent of unaccompanied females and 45 percent of unaccompanied males. At the same time, 7 percent of persons in families stayed at an emergency shelter every night (181 nights) during the study period compared to less than one percent of unaccompanied individuals. The median length of stay in emergency shelter for persons in families was 37 days compared to 17 days for unaccompanied females and 12 days for unaccompanied males.

There are also differences in length of stay between individuals and households with children who were served in transitional housing during the six-month reporting period for the second AHAR. The median length of stay for persons in households with children was 135 days compared to 94 days for unaccompanied females and 72 days for unaccompanied males. Transitional housing programs usually allow clients to stay for up to two years while working toward a permanent housing solution, so it not surprising that many people stayed there during the entire six-month (or 181 days) reporting period. Almost one-third (30 percent) of persons in families stayed in transitional housing the entire study period compared to only 13 percent of unaccompanied males. The share of unaccompanied females staying the whole period was close to that of persons in families (27 percent).

Looking Ahead - Participating communities have made much progress since the start of the AHAR 2 data collection period in early 2006, but additional work is needed to increase the precision of the estimates and the breadth of information reported. HUD is continuing outreach and technical assistance activities to help communities increase the number of providers participating in HMIS and improve the quality and usefulness of data for local needs. These efforts will also enable more communities to participate in AHAR. Simultaneously, HUD continues to provide technical assistance to communities on conducting one-night street and shelter counts, which will continue to be the source of information on the unsheltered homeless population in future AHAR reports.

HMIS implementation has progressed to the point that communities should now be able to provide data for an entire year. Since the third AHAR will cover a one-year period (October 1, 2006 through September 30, 2007), HMIS data will also be able to provide more detailed patterns of service use for people experiencing homelessness. This will help clarify the picture of current homeless service use and the needs of people experiencing homelessness.


Editor Note: Added the 1st report for review

The Annual Homeless Assessment Report to Congress(February 2007, 98 pages)