GAO-10-702, Homelessness: A Common Vocabulary Could Help Agencies Collaborate and Collect More Consistent Data


This is the accessible text file for GAO report number GAO-10-702 
entitled 'Homelessness: A Common Vocabulary Could Help Agencies 
Collaborate and Collect More Consistent Data' which was released on 
June 30, 2010. 

This text file was formatted by the U.S. Government Accountability 
Office (GAO) to be accessible to users with visual impairments, as 
part of a longer term project to improve GAO products' accessibility. 
Every attempt has been made to maintain the structural and data 
integrity of the original printed product. Accessibility features, 
such as text descriptions of tables, consecutively numbered footnotes 
placed at the end of the file, and the text of agency comment letters, 
are provided but may not exactly duplicate the presentation or format 
of the printed version. The portable document format (PDF) file is an 
exact electronic replica of the printed version. We welcome your 
feedback. Please E-mail your comments regarding the contents or 
accessibility features of this document to Webmaster@gao.gov. 

This is a work of the U.S. government and is not subject to copyright 
protection in the United States. It may be reproduced and distributed 
in its entirety without further permission from GAO. Because this work 
may contain copyrighted images or other material, permission from the 
copyright holder may be necessary if you wish to reproduce this 
material separately. 

Report to Congressional Requesters: 

United States Government Accountability Office: 
GAO: 

June 2010: 

Homelessness: 

A Common Vocabulary Could Help Agencies Collaborate and Collect More 
Consistent Data: 

GAO-10-702: 

GAO Highlights: 

Highlights of GAO-10-702, a report to congressional requesters. 

Why GAO Did This Study: 

Multiple federal programs provide homelessness assistance through 
programs targeted to those experiencing homelessness or through 
mainstream programs that broadly assist low-income populations. 
Programs’ definitions of homelessness range from including primarily 
people in homeless shelters or on the street to also including those 
living with others because of economic hardship. GAO was asked to 
address (1) the availability, completeness, and usefulness of federal 
data on homelessness, (2) the extent to which research identifies 
factors associated with experiencing homelessness, and (3) how 
differences in definitions and other factors impact the effectiveness 
of programs serving those experiencing homelessness. GAO reviewed 
laws, agency regulations, performance and planning documents, and data 
as well as literature on homelessness, and spoke with stakeholders, 
such as government officials and service providers, about potential 
barriers. 

What GAO Found: 

Federal agencies, including the Departments of Education (Education), 
Health and Human Services (HHS), and Housing and Urban Development 
(HUD), collect data on homelessness. However, these data are 
incomplete, do not track certain demographic information well over 
time, and are not always timely. HUD collects data and estimates the 
number of people who are homeless on a given night during the year and 
the number who use shelters over the course of the year; these 
estimates include the people who meet the definition of homelessness 
for HUD’s programs, but do not include all of those who meet broader 
definitions of homelessness used by some other agencies’ programs. For 
example, HUD’s counts would not include families living with others as 
a result of economic hardship, who are considered homeless by 
Education. Data from federally-funded mainstream programs such as 
HHS’s Temporary Assistance for Needy Families could improve agencies’ 
understanding of homelessness, but these programs have not 
consistently collected or analyzed information on housing status 
because this is not their primary purpose. 

Because research studies GAO reviewed often used different definitions 
of homelessness, relied on data collected at a point-in-time, and 
focused narrowly on unique populations over limited geographical 
areas, the studies cannot be compared or compiled to further an 
understanding of which factors are associated with experiencing 
homelessness. Furthermore, although researchers GAO interviewed and 
most studies noted the importance of structural factors such as area 
poverty rates, and those that analyzed these factors found them to be 
important, few studies considered them. Most of the studies analyzed 
only the association of individual-level factors such as demographic 
characteristics, but these studies often did not consider the same 
individual-level factors or agree on their importance. 

Many of the government officials, service providers, advocates, and 
researchers GAO interviewed stated that narrow or multiple definitions 
of homelessness have posed challenges to providing services for those 
experiencing homelessness, and some said that having different 
definitions made collaborating more difficult. For example, some said 
that persons in need of services might not be eligible for programs 
under narrower definitions of homelessness or might not receive 
services for which they were eligible because of confusion created by 
multiple definitions. Different definitions of homelessness and 
different terminology to address homelessness have made it difficult 
for communities to plan strategically for housing needs and for 
federal agencies such as Education, HHS, and HUD to collaborate 
effectively to provide comprehensive services. As long as agencies use 
differing terms to address issues related to homelessness, their 
efforts to collaborate will be impeded, and this in turn will limit 
the development of more efficient and effective programs. Commenting 
on a draft of this report, HHS and HUD raised concerns about its 
treatment of homelessness data. We characterize and respond to those 
comments within the report. 

What GAO Recommends: 

GAO recommends that Education, HHS, and HUD (1) develop a common 
vocabulary for homelessness; and (2) determine if the benefits of 
collecting data on housing status in targeted and mainstream programs 
would exceed the costs. To the extent that the agencies explicitly 
addressed the recommendations in their comments, they agreed with them. 

View [hyperlink, http://www.gao.gov/products/GAO-10-702] or key 
components. For more information, contact Alicia Cackley at (202) 512-
8678 or cackleya@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

Shortcomings in Federal Data Have Limited the Understanding of the 
Extent and Nature of Homelessness: 

Definitional Differences and Data and Methodological Issues in 
Research Studies Hinder Development of Comprehensive Understanding of 
Factors Associated with Homelessness: 

Definitional Issues Have Posed Challenges for Service Providers and 
Make Collaborating at Local and Federal Levels More Difficult: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope and Methodology: 

Appendix II: Bibliography of Studies GAO Reviewed That Analyze Factors 
Associated with Homelessness: 

Appendix III: Comments from the Department of Education: 

Appendix IV: Comments from the Department of Health and Human Services: 

Appendix V: Comments from the Department of Housing and Urban 
Development: 

Appendix VI: Comments from the Executive Director of the U. S. 
Interagency Council on Homelessness: 

Appendix VII: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: Targeted Homeless Programs: Purposes, Homelessness 
Definitions, and 2009 Funding Levels: 

Table 2: Selected Mainstream Programs That May Provide Benefits to 
Those Experiencing Homelessness: Purposes, Homelessness Definitions, 
and 2009 Funding Levels: 

Table 3: Federal Homelessness Data: Information on Housing Status, 
Required Frequency of Data Submission, and Level of Data Collected by 
Federal Agencies: 

Figures: 

Figure 1: Point-in-Time Count of Homeless Individuals and Persons in 
Families, 2005-2008: 

Figure 2: Homeless Students Served by the Education of Homeless 
Children and Youth Program, by Primary Nighttime Residence Type and 
School Year: 

Abbreviations: 

Continuum: Continuum of Care: 

DOJ: Department of Justice: 

DOL: Department of Labor: 

Education: Department of Education: 

HEARTH Act: Homeless Emergency Assistance and Rapid Transition to 
Housing Act: 

HMIS: Homelessness Management Information Systems: 

HHS: Department of Health and Human Services: 

HUD: Department of Housing and Urban Development: 

HUD-VASH: Housing and Urban Development-Veterans Affairs Supportive 
Housing: 

HVRP: Homeless Veterans' Reintegration Program: 

Interagency Council: Interagency Council on Homelessness: 

McKinney-Vento Ac:t McKinney-Vento Homeless Assistance Act: 

PATH: Projects for Assistance in Transition from Homelessness: 

PHA: public housing authority: 

Recovery Act: American Recovery and Reinvestment Act of 2009: 

RHYMIS: Runaway and Homeless Youth Management Information System: 

SNAP: Supplemental Nutrition Assistance Program: 

SOAR: SSI/SSDI Outreach, Access and Recovery Initiative: 

SSA: Social Security Administration: 

SSI: Supplemental Security Income: 

SSDI: Supplemental Security Disability Insurance: 

TANF: Temporary Assistance for Needy Families: 

USDA: U.S. Department of Agriculture: 

VA: Department of Veterans Affairs: 

WIA: Workforce Investment Act: 

[End of section] 

United States Government Accountability Office: 
Washington, DC 20548: 

June 30, 2010: 

[See PDF for image] 

[End of figure] 

Congressional Requesters: 

For some time, the federal government and the nation at large have 
been concerned about the number of people who are experiencing 
homelessness, but approaches to solving the problem have varied over 
time. In part because of the financial crisis that began in 2007, 
concerns about homelessness have come to the fore again. Multiple 
federal agencies administer programs designed to address the needs of 
those experiencing homelessness, but the programs use different 
definitions of homelessness to determine eligibility. These 
differences in definitions were an important part of the discussions 
leading to the enactment in 2009 of the Homeless Emergency Assistance 
and Rapid Transition to Housing Act (HEARTH Act).[Footnote 1] The 
HEARTH Act reauthorized the McKinney-Vento Homeless Assistance Act 
(McKinney-Vento Act), which contained different definitions of 
homelessness for the federal programs it authorized. For programs 
administered by the Department of Housing and Urban Development (HUD), 
the McKinney-Vento Act generally defined homelessness as being on the 
street or in an emergency or transitional shelter, while for programs 
administered by the Department of Education (Education), the McKinney-
Vento Act defined homelessness more broadly to include homeless 
children and youths who are sharing the housing of other persons due 
to loss of housing or economic hardship and those living in motels or 
camping grounds due to the lack of adequate alternative 
accommodations. Any discussion of homeless definitions is further 
complicated by programs using other definitions. For example, the 
Runaway and Homeless Youth Act authorizes programs administered by the 
Department of Health and Human Services (HHS) that define homelessness 
as lacking a safe living arrangement. In part because of these 
differences in definitions, concerns also have been raised about 
whether existing data on homelessness provide a complete picture of 
its extent and nature. 

Previously, we reported on agency coordination and evaluation of 
homelessness programs and the barriers faced by those experiencing 
homelessness in using mainstream programs.[Footnote 2] In response to 
your request, this report updates some of that earlier information and 
specifically addresses (1) the availability, completeness, and 
usefulness of data on homelessness collected by federal programs; (2) 
the extent to which research identifies factors associated with 
homelessness; and (3) how differences in the definitions of 
homelessness and other factors, such as the level of agency 
collaboration, may impact the effectiveness of programs serving those 
experiencing homelessness. 

To address these objectives, we reviewed relevant laws, regulations, 
and government reports across a number of programs specifically 
targeted to address issues related to homelessness as well as 
mainstream programs--programs such as Temporary Assistance for Needy 
Families (TANF), Head Start, and Public Housing--that are generally 
designed to help low-income individuals and families achieve or retain 
their economic self-sufficiency and often provide services to people 
experiencing homelessness. We also interviewed officials at HUD, HHS, 
and Education; the U.S. Interagency Council on Homelessness 
(Interagency Council); and the Departments of Justice (DOJ) and Labor 
(DOL). We conducted in-depth interviews with a variety of 
stakeholders, including advocates and researchers, as well as service 
providers, state and local government officials, and HUD field staff 
that had extensive experience with homeless programs. To gather 
perspectives on our objectives, we conducted four site visits to large 
and medium-sized urban areas that were geographically distributed 
across the United States. These areas were in California, Illinois, 
Massachusetts, and South Carolina. To determine the availability, 
completeness, and usefulness of data, we reviewed the methodologies 
and reliability of program data collected by HUD, HHS, and Education 
and interviewed stakeholders about their use of the data. We also 
analyzed estimates of the extent of homelessness that were derived 
from federal data systems such as the American Community Survey. To 
determine the extent to which the research identifies factors 
associated with homelessness, we reviewed studies published from 1998 
to 2009; selected 45 studies that we judged to be sufficiently 
rigorous; and systematically reviewed their methodologies, findings, 
and limitations. We chose 1998 as a starting point because welfare 
reform--which impacted some homeless families--had been implemented by 
that date and may have affected research findings. To determine what 
factors create barriers to serving those experiencing homelessness, we 
developed a list of potential barriers from the literature and 
discussions with researchers and asked government officials, service 
providers, advocates, and researchers who represented a range of 
activities and views related to homelessness to select the three most 
important barriers from that list. We determined the relative 
importance of the barriers chosen by summing the number of times an 
item was selected. Finally, we asked these stakeholders to describe 
how federal agencies collaborate with regard to homelessness and 
reviewed joint program and agency planning and performance documents. 
See appendix I for more detailed information on our scope and 
methodology. 

We conducted this performance audit from May 2009 to June 2010 in 
accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

Background: 

Definitions of Homelessness: 

Congress first provided a general definition of homeless individuals 
in 1987 in what is now called the McKinney-Vento Act.[Footnote 3] In 
2002, Congress added a definition for homeless children and youths to 
be used in educational programs.[Footnote 4] Prior to the enactment of 
the HEARTH Act, the McKinney-Vento Act generally defined a homeless 
individual (McKinney-Vento Individual) as someone who lacks a fixed, 
regular, and adequate nighttime residence and has a nighttime 
residence that is a supervised shelter designed to provide temporary 
accommodations; an institution providing a temporary residence for 
individuals awaiting institutionalization; or a place not designed 
for, nor ordinarily used as, a regular sleeping accommodation. 
[Footnote 5] However, in the provisions on education of children and 
youths, the McKinney-Vento Act also specifically included children and 
youths who are sharing the housing of other persons due to loss of 
housing, economic hardship, or a similar reason (that is, are doubled 
up); living in motels, hotels, trailer parks, or camping grounds due 
to the lack of alternative adequate accommodations; awaiting foster 
care placement; or living in substandard housing (McKinney-Vento 
Children and Youth).[Footnote 6] 

For its homeless assistance programs, HUD has interpreted the McKinney-
Vento Act definitions so that a homeless individual is someone who 
resides in places not meant for human habitation, such as in cars, 
abandoned buildings, housing that has been condemned by housing 
officials, or on the street, in an emergency shelter or transitional 
or supportive housing, or any of these places, but is spending a short 
time (up to 30 consecutive days) in a hospital or other institution. 
Additionally, individuals are considered homeless if they are being 
evicted within a week from a private dwelling and no subsequent 
residence has been identified and the person lacks the resources and 
support networks needed to obtain housing; discharged within a week 
from an institution in which the person has been a resident for 30 or 
more consecutive days and no subsequent residence has been identified; 
or fleeing a domestic violence situation.[Footnote 7] 

The HEARTH Act includes changes in the general definition of 
homelessness, but the new definition and associated regulations had 
not taken effect by June 2010. The HEARTH Act broadened the general 
definition and provided greater statutory specificity concerning those 
who should be considered homeless but did not change the McKinney-
Vento Children and Youth definition. For example, the HEARTH Act 
definition includes individuals and families that will be evicted in, 
or who can otherwise demonstrate that they will not be able to remain 
in their current living place for more than, 2 weeks.[Footnote 8] The 
HEARTH Act definition includes some individuals, families, and youths 
who would have been considered homeless under the McKinney-Vento 
Children and Youth definition but not under the prior individual 
definition.[Footnote 9] 

Some federal programs that were authorized outside of the McKinney- 
Vento Act use other definitions of homelessness. For example, the 
Runaway and Homeless Youth Act, first introduced as the Runaway Youth 
Act of 1974, defined a homeless youth as being generally from the ages 
of 16 to 22, unable to live in a safe environment with a relative, and 
lacking any safe alternative living arrangements.[Footnote 10] 

Within various programs, the definition of homelessness determines 
whether individuals are eligible for program benefits. For the 
Education of Homeless Children and Youth program, meeting the 
definition entitles the student to certain benefits; however, in other 
cases, such as HUD's homeless assistance programs or HHS's Runaway and 
Homelessness Youth programs, benefits are limited by the amount of 
funds appropriated for the program. For these programs, meeting the 
definition of homelessness does not necessarily entitle individuals or 
families to benefits. In addition, programs have other eligibility 
criteria, such as certain income levels, ages, or disability status. 

Numerous Federal Programs Address Homelessness: 

As illustrated in table 1, programs that provide targeted assistance 
primarily to those experiencing homelessness have different purposes, 
definitions of homelessness, and funding levels. One of these 
programs, HUD's Homeless Prevention and Rapid Rehousing Program, was 
created under the American Recovery and Reinvestment Act (Recovery 
Act) of 2009, and many others received additional funding under that 
act. 

Table 1: Targeted Homeless Programs: Purposes, Homelessness 
Definitions, and 2009 Funding Levels:
(Thousands of dollars): 

Federal agency: Education; 
Program: Education for Homeless Children and Youth; 
Purpose: Ensure that homeless children and youths have equal access to 
free and appropriate public education and to facilitate their 
enrollment, attendance, and success in school; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $65,427; 
Funding FY 2009 and additional Recovery Act: Recovery Act: $70,000. 

Federal agency: Department of Homeland Security/Federal Emergency 
Management Agency; 
Program: Emergency Food and Shelter; 
Purpose: Supplement and expand ongoing efforts to provide shelter, 
food and supportive services for homeless and hungry individuals 
nationwide; 
Homelessness definition[A]: McKinney-Vento Children and Youth; 
Funding FY 2009 and additional Recovery Act: FY 2009: $200,000; 
Funding FY 2009 and additional Recovery Act: Recovery Act: $100,000. 

Federal agency: DOL; 
Program: Homeless Veterans' Reintegration Program; 
Purpose: Provide services to assist in reintegrating homeless veterans 
into meaningful employment and stimulate the development of effective 
service delivery systems to address problems facing homeless veterans; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $26,330; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: DOJ; 
Program: Transitional Housing Assistance for Victims of Domestic 
Violence, Stalking, or Sexual Assault; 
Purpose: Provide housing assistance to those fleeing domestic 
violence, dating violence, sexual assault, or stalking; 
and for whom emergency shelter services or other crisis intervention 
services are not sufficient; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $18,000; 
Funding FY 2009 and additional Recovery Act: Recovery Act: $50,000. 

Federal agency: HHS; 
Program: Healthcare for the Homeless; 
Purpose: Recognize complex needs of homeless persons and strive to 
provide a coordinated, comprehensive approach to health care, 
including substance abuse and mental health services; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $188,342; 
Funding FY 2009 and additional Recovery Act: Recovery Act: $174,000. 

Federal agency: HHS; 
Program: Projects for Assistance in Transition from Homelessness 
(PATH); 
Purpose: Provide community-based mental health, substance abuse, and 
other support, including limited housing to individuals with serious 
mental illness who are experiencing or at risk of homelessness; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $59,687; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HHS; 
Program: Grants for the Benefit of Homeless Individuals; 
Purpose: Enables communities to expand and strengthen their treatment 
services for homeless persons with substance abuse and mental health 
disorders and to link these services to stable housing; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $42,879; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HHS; 
Program: Services in Supportive Housing; 
Purpose: Provide treatment and support services to people experiencing 
homelessness and severe mental illness or co-occurring mental and 
substance use disorders in coordination with permanent supportive 
housing programs and resources; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $34,556; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HHS; 
Program: Runaway and Homeless Youth - Basic Services; 
Purpose: Meet immediate needs of runaway and homeless youths and their 
families and provide temporary emergency shelter for youths under age 
20 and some other services for those under 18; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $53,469; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HHS; 
Program: Runaway and Homeless Youth - Transitional Living Program for 
Older Homeless Youth; 
Purpose: Provide long-term residential services to homeless youths 
aged 16-21 to help them transition to self-sufficiency; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $43,765; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HHS; 
Program: Runaway and Homeless Youth -Street Outreach Program; 
Purpose: Help young people get off the streets to prevent sexual abuse 
and exploitation; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $17,721; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HUD; 
Program: Homeless Assistance Programs; 
Purpose: Address homelessness through the provision of emergency 
shelter, supportive services, transitional housing, permanent housing, 
and prevention resources to assist individuals in shelters or on the 
streets attain permanent housing and self-sufficiency; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $1,677,000; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HUD; 
Program: Homeless Prevention and Rapid Re-housing; 
Purpose: Provide homelessness prevention assistance to households that 
would otherwise become homeless--many due to the economic crisis--and 
provide assistance to rapidly re-house persons who are homeless; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: [Empty]; 
Funding FY 2009 and additional Recovery Act: $1,500,000. 

Federal agency: Department of Veterans Affairs (VA)[B]; 
Program: Healthcare for Homeless Veterans; 
Purpose: Perform outreach to identify homeless veterans eligible for 
VA services and assist them in accessing appropriate healthcare and 
benefits; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $80,219; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: Department of Veterans Affairs (VA)[B]; 
Program: Homeless Providers Grants and Per Diem Program; 
Purpose: Promote the development and provision of supportive housing 
and supportive services to help homeless veterans achieve residential 
stability, increase skill levels, and obtain greater self- 
determination; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $130,000; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: Department of Veterans Affairs (VA)[B]; 
Program: Domiciliary Care for Homeless Veterans; 
Purpose: Provide services to economically disadvantaged veterans; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $98,789; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: Department of Veterans Affairs (VA)[B]; 
Program: Compensated Work Therapy Transitional Residence Program; 
Purpose: Provide vocational opportunities in residential setting for 
veterans recovering from chronic mental illness, chemical dependency, 
and homelessness; 
Homelessness definition[A]: Other; 
Funding FY 2009 and additional Recovery Act: FY 2009: $22,206; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: Department of Veterans Affairs (VA)[B]; 
Program: Loan Guarantee for Transitional Housing for Homeless Veterans; 
Purpose: Increase the amount of housing available, and provide 
services to encourage addiction recovery; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $45; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: Department of Veterans Affairs (VA)[B]; 
Program: Supportive Services for Veteran Families; 
Purpose: Provide supportive services to very low-income veteran 
families in or transitioning to permanent housing; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $218; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Federal agency: HUD-VA; 
Program: HUD-and VA Supported Housing (HUD-VASH); 
Purpose: Provide subsidized housing and services for homeless veterans; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding FY 2009 and additional Recovery Act: FY 2009: $54,128; 
Funding FY 2009 and additional Recovery Act: Recovery Act: [Empty]. 

Source: Congressional Research Service, Homelessness: Targeted Federal 
Programs and Recent Legislation, RL30442 (Washington, D.C.: Jan. 20, 
2010); GAO; and listed agencies. 

[A] Definition classifications depend on statutory definitions, 
program regulations, or statements by agency officials concerning 
their programs' definitions of homelessness. Individuals and families 
not defined as homeless under a program may in some specified 
circumstances be eligible for program services. 

[B] Funding for VA programs includes estimated program obligations 
while funding for other programs includes fiscal year 2009 and 
additional Recovery Act appropriated amounts. 

[End of table] 

HUD's Homeless Assistance Programs comprise a number of individual 
programs. These include the Emergency Shelter Grant Program, under 
which funding is provided on a formula basis, and competitive programs 
funded under the umbrella of the Continuum of Care (Continuum). The 
latter include the Supportive Housing, Shelter Plus Care, and Single 
Room Occupancy Programs. A Continuum is a group of providers in a 
geographical area that join to provide homeless services and apply for 
these grants. The Continuum is also responsible for planning homeless 
services, setting local priorities, and collecting homelessness data. 

Additionally, many federally-funded mainstream programs provide 
services for which those experiencing homelessness may be eligible. 
Some of these programs are required to provide services to those 
experiencing homelessness and may define it, others allow local 
providers to choose to target certain services to those experiencing 
homelessness or provide homelessness preferences using locally 
determined definitions, and still other programs do not distinguish 
between those experiencing homelessness and those not experiencing it 
(see table 2). 

Table 2: Selected Mainstream Programs That May Provide Benefits to 
Those Experiencing Homelessness: Purposes, Homelessness Definitions, 
and 2009 Funding Levels:
(Thousands of dollars): 

Federal Agency: DOL; 
Program: Workforce Investment Act (WIA) Youth and Adult Programs; 
Purpose: Provide employment assistance to those who face barriers, 
such as homelessness; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding[B] FY 2009: $1,785,609[C]. 

Federal Agency: DOL; 
Program: Job Corps; 
Purpose: Assist eligible youth in developing into responsible, 
employable, and productive citizens and successful members of the 
workforce; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding[B] FY 2009: $1,683,938. 

Federal Agency: HHS; 
Program: Children's Health Insurance Program; 
Purpose: Provide health insurance to children in families with very 
low incomes; 
Homelessness definition[A]: McKinney-Vento Children and Youth; 
Funding[B] FY 2009: $10,600,000. 

Federal Agency: HHS; 
Program: Community Health Centers; 
Purpose: Serve populations with limited access to health care, 
including individuals and families experiencing homelessness; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $2,150,000. 

Federal Agency: HHS; 
Program: Community Mental Health Services Block Grant; 
Purpose: Provide and encourage the development of creative and cost-
effective community-based care for people with serious mental 
disorders; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $420,774. 

Federal Agency: HHS; 
Program: Head Start; 
Purpose: Promote school readiness by enhancing the social and 
cognitive development of children through the provision of 
educational, health, nutritional, social and other services to 
enrolled children and families; 
Homelessness definition[A]: McKinney-Vento Children and Youth; 
Funding[B] FY 2009: $7,100,000. 

Federal Agency: HHS; 
Program: John H. Chafee Foster Care Independence Program; 
Purpose: Assist current and former foster care youths achieve self-
sufficiency; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $140,000. 

Federal Agency: HHS; 
Program: Medicaid; 
Purpose: Provide access to medical services for individuals and 
families who meet certain state requirements such as having a low 
income; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $216,600,000. 

Federal Agency: HHS; 
Program: Ryan White HIV/AIDS Program; 
Purpose: Improve the quality and availability of care for low-income 
uninsured and underinsured individuals and families affected by HIV 
disease; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding[B] FY 2009: $2,240,000. 

Federal Agency: HHS; 
Program: Substance Abuse Prevention and Treatment Block Grant; 
Purpose: Provide substance abuse treatment and prevention services to 
individuals in communities at risk for substance abuse/dependence; 
Homelessness definition[A]: Other; 
Funding[B] FY 2009: $1,779,000. 

Federal Agency: HHS; 
Program: Temporary Assistance for Needy Families (TANF); 
Purpose: Provide temporary cash assistance and services for low-income 
families with children; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $16,500,000. 

Federal Agency: HUD; 
Program: Public Housing; 
Purpose: Provide decent and safe rental housing for eligible low-
income families, the elderly, and persons with disabilities; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $10,905,000. 

Federal Agency: HUD; 
Program: Section 8 Tenant-Based Assistance; 
Purpose: Assist very low income families, the elderly, and persons 
with disabilities to afford decent, safe, and sanitary housing in the 
private market; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $16,225,000. 

Federal Agency: HUD; 
Program: HOME Investment Partnership Program; 
Purpose: Expand the supply of affordable housing and increase the 
capacity of state and local governments and nonprofit organizations in 
developing such housing; 
Homelessness definition[A]: Permit local determination; 
Funding[B] FY 2009: $1,825,000. 

Federal Agency: HUD; 
Program: Community Development Block Grant; 
Purpose: Address a wide range of unique community development needs; 
Homelessness definition[A]: McKinney-Vento-Individual; 
Funding[B] FY 2009: $3,899,999. 

Federal Agency: Social Security Administration (SSA); 
Program: Supplemental Security Income (SSI); 
Purpose: Provide cash benefits to people with disabilities who have 
limited income, assets, and work history; 
Homelessness definition[A]: Other; 
Funding[B] FY 2009: $48,000,000. 

Federal Agency: Social Security Administration (SSA); 
Program: Supplemental Security Disability Insurance (SSDI); 
Purpose: Provide cash benefits related to prior earnings to people 
with disabilities that have a Social Security work record; 
Homelessness definition[A]: Other; 
Funding[B] FY 2009: $118,114,000. 

Federal Agency: United States Department of Agriculture (USDA); 
Program: Supplemental Nutrition Assistance Program (SNAP); 
Purpose: Provide cash-like benefits to low-income households that can 
be used to purchase food from participating retail stores; 
Homelessness definition[A]: Other; 
Funding[B] FY 2009: $53,800,000. 

Federal Agency: United States Department of Agriculture (USDA); 
Program: Special Supplemental Nutrition Program for Women, Infants and 
Children; 
Purpose: Protect the health of those at nutritional risk by providing 
nutritious foods, healthy eating information, and health-care 
referrals; 
Homelessness definition[A]: Other; 
Funding[B] FY 2009: $6,860,000. 

Source: GAO, DOL, HHS, and HUD. 

[A] Definition classifications depend on statutory definitions that 
refer explicitly to a McKinney-Vento definition or are similar to 
those statutory definitions, to program regulations, or statements by 
agencies. The other category includes both definitions that don't fall 
into the other listed categories as well as those programs that have 
no references to definitions of homelessness. 

[B] Funding does not include Recovery Act funding. 

[C] WIA funding is appropriated for a program year which runs from 
July 1, 2009 to June 30, 2010. 

[End of table] 

The McKinney-Vento Act also authorized the creation of the U. S. 
Interagency Council on Homelessness (Interagency Council).[Footnote 
11] Initially, the main functions of the Interagency Council revolved 
around using public resources and programs in a more coordinated 
manner to meet the needs of those experiencing homelessness. The 
McKinney-Vento Act specifically mandated the council to identify 
duplication in federal programs and provide assistance to states, 
local governments, and other public and private nonprofit 
organizations to enable them to serve those experiencing homelessness 
more effectively. In the HEARTH Act, the council, which includes 19 
agencies, was given the mission of coordinating the federal response 
to homelessness and creating a national partnership at every level of 
government and with the private sector to reduce and end homelessness. 
[Footnote 12] This act also mandates that the Interagency Council 
develop and annually update a strategic plan to end homelessness. 

Some Agencies Collect Data on Those Experiencing Homelessness: 

Several agencies overseeing targeted homelessness programs are 
required to collect data on segments of the homeless population. As 
illustrated in table 3, HUD, HHS, and Education all have met their 
requirements through their own data collection and these sources 
differ in housing data collected and level of aggregation.[Footnote 
13] In addition, the data collected necessarily reflect the 
definitions of homelessness included in the statutes that govern the 
relevant programs. 

Table 3: Federal Homelessness Data: Information on Housing Status, 
Required Frequency of Data Submission, and Level of Data Collected by 
Federal Agencies: 

Department: HUD; 
Data source: Point-in-Time Count; 
Housing status data collected: Sheltered and unsheltered on night of 
count; 
Required frequency of data submission: Biennial[A]; 
Level of data collected: Aggregate. 

Department: HUD; 
Data source: Homeless Management Information System (HMIS)[B]; 
Housing status data collected: Housing status, type of residence on 
night before program entry (shelter, hospital, doubled up, etc.), 
length of stay at last residence, and zip code of last permanent 
address; 
Required frequency of data submission: Annual; 
Level of data collected: Aggregate. 

Department: HHS; 
Data source: Runaway and Homeless Youth Management Information System 
(RHYMIS); 
Housing status data collected: Living situation (shelter, street, 
private residence, correctional facility, etc.) at program entry and 
exit; 
Required frequency of data submission: Semiannual; 
Level of data collected: Individual. 

Department: Education; 
Data source: Consolidated State Performance Reports; 
Housing status data collected: Primary nighttime residence of students 
in homeless education program (street, shelter, doubled up, etc.); 
Required frequency of data submission: Annual; 
Level of data collected: Aggregate. 

Source: GAO analysis of HUD, HHS, and Education documents. 

Note: HUD and Education data are collected locally at the individual 
level, and they are reported in aggregate to the agency. 

[A] HUD requires Continuums to conduct a biennial count, but some 
Continuums conduct point-in-time counts every year. In 2008--a year 
that HUD did not require a count--approximately 67 percent of 
communities submitted point-in-time count data. 

[B] In June 2009, HUD revised its HMIS data standards to include 
"housing status" as a universal data element for HMIS. This element 
tracks whether the client is homeless, at imminent risk of 
homelessness, at risk of homelessness, or stably housed. 

[End of table] 

* Under the McKinney-Vento Act, HUD is to develop an estimate of 
homeless persons in sheltered and unsheltered locations at a 1-day 
point in time, so HUD requires Continuums to conduct a count of the 
sheltered and unsheltered homeless in their jurisdictions in January 
of every other year.[Footnote 14] Additionally, pursuant to the 2001 
amendments to the McKinney-Vento Act, HUD was to develop a system to 
collect and analyze data on the extent of homelessness and the 
effectiveness of McKinney-Vento Act programs.[Footnote 15] As a 
result, HUD developed a set of technical data standards for the 
Homelessness Management Information System (HMIS), which sets minimum 
privacy, security, and technical requirements for local data 
collection on the characteristics of individuals and families 
receiving homelessness services. HMIS data standards allowed 
communities to continue using locally developed data systems and adapt 
them to meet HUD standards. Local Continuums are responsible for 
implementing HMIS in their communities, and Continuums can choose from 
many HMIS systems that meet HUD's data standards. HUD officials said 
that by allowing Continuums to choose from multiple systems, more 
service providers participate and Continuums and service providers can 
modify existing systems to meet HUD standards and the community's 
goals. Continuums report aggregated data to HUD annually. Results from 
analysis of the point-in-time count and HMIS data are reported in 
HUD's Annual Homelessness Assessment Report to Congress. 

* Pursuant to the Runaway and Homeless Youth Act, HHS requires all 
service providers to collect data on youths who receive services 
through the Runaway and Homeless Youth Program. Grantees submit these 
data every 6 months to the Runaway and Homeless Youth Management 
Information System (RHYMIS), a national database that includes 
unidentified individual-level data. 

* To demonstrate compliance with the Elementary and Secondary 
Education Act of 1965, Education requires states to complete 
Consolidated State Performance Reports that include data on homeless 
children and youths being served by Elementary and Secondary Education 
Act programs and the Education of Homeless Children and Youth Program, 
as amended. The McKinney-Vento Act requires local school districts to 
have homelessness liaisons, who work with other school personnel and 
those in the community to identify homeless children and youths, 
provide appropriate services and support, and collect and report data. 
States aggregate local data and report to Education annually 
cumulative numbers of homeless students enrolled in public schools by 
grade and primary nighttime residence. 

As part of its decennial population and housing census, the U.S. 
Census Bureau has programs designed to count people experiencing 
homelessness. The Census counts people at places where they receive 
services (such as soup kitchens or domestic violence shelters), as 
well as at targeted nonshelter outdoor locations. While the Census 
makes an effort to count all residents, including those experiencing 
homelessness, the 2010 Census does not plan to report a separate count 
of the population experiencing homelessness or a count of the 
population who use the services.[Footnote 16] 

Shortcomings in Federal Data Have Limited the Understanding of the 
Extent and Nature of Homelessness: 

Homelessness Data of Federal Agencies Have Shortcomings: 

Although federal agencies collect data on those experiencing 
homelessness, these data have a number of shortcomings and 
consequently do not capture the true extent and nature of 
homelessness. Some of these shortcomings derive from the difficulty of 
counting a transient population that changes over time, lack of 
comprehensive data collection requirements, and the time needed for 
data analysis. As a result of these shortcomings, the data have 
limited usefulness. Complete and accurate data are essential for 
understanding and meeting the needs of those who are experiencing 
homelessness and to prevent homelessness from occurring. According to 
HUD, communities need accurate data to determine the extent and nature 
of homelessness at a local level, plan services and programs to 
address local needs, and measure progress in addressing homelessness. 
HUD needs accurate data to fulfill its reporting obligations to 
Congress and to better understand the extent of homelessness, who it 
affects, and how it can best be addressed. 

HUD's point-in-time count is the only data collection effort designed 
to obtain a national count of those experiencing homelessness under 
the McKinney-Vento Individual definition, and approximately 450 
Continuums conduct a point-in-time count in January of every other 
year. However, service providers and researchers we interviewed 
expressed concerns about the ability of HUD's point-in-time count to 
fully capture many of those experiencing homelessness for reasons 
including the following: 

* People experiencing homelessness are inherently difficult to count. 
They are mobile, can seek shelter in secluded areas, and may not wish 
to attract the notice of local government officials. 

* Point-in-time counts do not recognize that individuals and families 
move in and out of homelessness and can experience it for varying 
lengths of time. These counts more likely count those experiencing 
homelessness for long periods rather than those experiencing it 
episodically or for short periods.[Footnote 17] 

* Although homelessness can be episodic, the count is done biennially 
in January, which might lead to an undercount of families because 
landlords and others may be reluctant to evict families when the 
weather is cold or school is ongoing. 

* Count methodologies vary by Continuum, can change from year to year, 
and might not be well implemented because counters are volunteers who 
may lack experience with the population. 

* Large communities do not necessarily attempt to count all of those 
experiencing homelessness but rather may use estimation procedures of 
varying reliability. 

HUD provides technical assistance to communities, which helps them to 
develop and implement point-in-time count methodologies, and HUD 
officials said that methodologies and the accuracy of the count have 
improved. Additionally, HUD officials said that as part of their 
quality control efforts, they contacted 213 Continuums last year to 
address errors or inconsistencies in their data from fiscal year 2008. 
A communitywide point-in-time count demands considerable local 
resources and planning, and communities rely on volunteers to conduct 
counts of the unsheltered population. Continuums do not receive any 
funding from HUD to conduct the point-in-time counts, and using 
professionals or paid staff to conduct the count could be costly. 

Other federal data collected on those experiencing homelessness 
primarily or only captures clients being served by federally-funded 
programs. As a result, federal data do not capture some people seeking 
homeless assistance through nongovernmental programs, or others who 
are eligible for services but are not enrolled. For instance, while 
HUD grantees are required to participate in HMIS, participation is 
optional for shelters that do not receive HUD funding. HUD can use the 
annual Continuum funding application to assess the extent to which 
those shelters not receiving HUD funding participate in HMIS. In their 
funding applications Continuums provide an inventory of shelter beds 
in their community and also provide the percentage of those beds that 
are located in shelters that participate in HMIS. HMIS participation 
rates vary widely across communities and shelter types. For example, 
one of the locations we visited reported data for less than 50 percent 
of its beds for transitional shelters, while another reported data for 
more than 75 percent of its beds. HUD officials said that while some 
Continuums have been slower to implement HMIS and receive full 
participation from their providers than others, according to HUD's 
2009 national housing inventory data of homeless programs, 75 percent 
of all shelter beds were covered by HMIS, including programs that do 
not receive HUD funding. The Violence Against Women Act prohibits 
service providers from entering individual-level data into HMIS for 
those in domestic violence shelters.[Footnote 18] Similarly, RHYMIS 
collects data on those clients using its residential systems, but 
these serve only a small percentage of the estimated number of youths 
experiencing homelessness. HHS officials stated that nationwide, they 
only fund approximately 200 transitional living centers for young 
adults. Education also does not fully capture the extent of 
homelessness among school-aged children because all of the districts 
we visited used a system of referrals and self-reporting to identify 
those children. In one of the school districts we visited, an official 
said that, based on estimates of the number of children experiencing 
homelessness under the McKinney-Vento Children and Youth definition, 
the district was serving about half of those students. Many of the 
school officials and advocates with whom we spoke said the term 
homelessness carried a stigma that made people reluctant to be so 
identified, and two school systems had removed the word from the name 
of their programs. 

Additionally, federal data systems on homelessness may count the same 
individual more than once. HUD designed HMIS to produce an 
unduplicated count--one that ensures individuals are counted only 
once--of those experiencing homelessness within each Continuum. 
Providers in the same Continuum use the same HMIS system and some 
Continuums have designed an open system, where providers can view all 
or part of an individual's record from another provider within the 
Continuum. This is useful to providers because it helps them to 
understand an individual or family's service needs. It also allows 
them to produce an unduplicated count of those using services in the 
Continuum because every person receiving services in the Continuum has 
a unique identifier in HMIS. However, it is difficult to share data 
across Continuums and this can be done only if Continuums have signed 
agreements that protect privacy and are using the same HMIS system. 
Thus, clients may be entered into HMIS in more than one Continuum and 
counted more than once. Nonetheless, some states have constructed 
statewide HMIS systems to help avoid duplication in the data. Because 
RHYMIS has individual data on all program recipients in a single 
database, HHS can obtain an accurate count on the number of youths 
served by its residential programs. Education data also may be 
duplicative. Because students generally are counted as homeless if 
they experience homelessness at any point during the school year, if 
they change school districts during the year, they could be counted as 
homeless in both systems. While each agency makes efforts to avoid 
duplication in its data, it is not possible to determine how many 
total unique individuals federal homelessness programs have served 
because HUD, HHS, and Education data systems generally do not 
interface or share data. 

Further, the data in HMIS may not always accurately reflect the 
demographic information on families and individuals seeking shelter. 
For example, HMIS provides data for individuals and families but the 
system may not accurately identify family members and track the 
composition of families over time. Using HMIS, service providers 
associate individuals entering into a shelter with a family if family 
members enter the shelter together. However, some families split up to 
obtain shelter, so the system would not track all families over time. 
In one of our site visit locations, we met a mother and son who were 
split up and placed in separate shelters. Because the mother was in an 
individual shelter, and the son was in a youth shelter, HMIS would not 
associate these two as a family. Further, one service provider we 
spoke with said that HMIS may not always accurately track demographic 
information on individuals seeking emergency shelter. 

Some researchers and advocates told us that HMIS's design limited its 
usefulness, and the extent to which service providers found that the 
HMIS system their Continuum had implemented was useful varied across 
the four locations we visited. For example, a researcher who has 
extensively used HMIS told us that if service providers used the data 
they collected for HMIS to manage their programs, they would implement 
processes to help ensure data quality. But in three of the four 
locations we visited, many providers said they were unable to use HMIS 
for program administration and client case management. Many providers 
noted that they often had to enter information in several different 
databases, and they generally used their own database to administer 
their programs. Additionally, we found only two providers who 
developed data export tools that allowed their private systems to 
upload data to HMIS, and in both cases, the providers were unable to 
use their new tools after the Continuum switched HMIS software. HHS 
officials told us that they support providers' development of tools to 
link data systems, but they do not provide funding for this endeavor. 
In contrast, service providers in one location we visited reported 
that the HMIS system they had adopted had options that allowed them to 
conduct comprehensive case management for clients and produce all of 
the reports required by the various organizations funding their 
programs and operations. HUD officials said that a community's success 
in using HMIS for program administration and client case management 
depends on a variety of factors including staff capability and the 
quality of the HMIS software that they chose to purchase or develop. 

HUD and Education data also have shortcomings and limited usefulness 
because of the time lag between initial data collection and the 
reporting of the data. HUD published the most recent report to 
Congress, which provided data for October 2008-September 2009, in June 
2010. Education expected to publish data for the 2008-2009 school year 
in June 2010. Because of the time lag in availability of HUD and 
Education data, they have limited usefulness for understanding current 
trends in homelessness and the ongoing effects of the recession. 
However, HUD officials report that they have made progress in reducing 
the time it takes to analyze data and publish its annual report to 
Congress. The time lag from data collection to report issuance has 
decreased from almost 2 years to less than 1 year, but collecting data 
on homelessness and producing national estimates takes time, and HUD 
officials said there will always be some time lag. Additionally, in 
recognition of these shortcomings, HUD recently introduced the 
Homeless Pulse Project, which collects quarterly homeless shelter data 
from nine communities. These communities volunteered to submit data on 
a more frequent basis, but they are not representative of Continuums 
nationwide. HUD plans to expand the Pulse Project and add 
approximately 30 Continuums that have volunteered to participate. HHS' 
RHYMIS data are more timely because grantees submit data every 6 
months, and HHS makes the data available online approximately 1 month 
after the end of the reporting period. 

Although a researcher with special expertise in HMIS and several 
advocates we interviewed cited some examples of incompleteness or 
inaccuracy in HMIS data, agencies and Continuums have been trying to 
improve the completeness and accuracy of their data. For example, HUD 
provides incentives for Continuums to increase HMIS participation 
rates. In its competitive grant process, HUD evaluates the level to 
which Continuums participate in HMIS. HUD officials have also provided 
technical assistance to Continuums to assist them in increasing local 
HMIS participation rates. HMIS rates have increased over time. Several 
Continuums we contacted have been conducting outreach to private and 
faith-based providers to encourage them to use HMIS to improve data on 
homelessness. Additionally, according to HUD officials, HMIS data have 
been used to conduct research on the prevalence of homelessness, the 
costs of homelessness, and the effectiveness of interventions to 
reduce homelessness. Further, HUD supplements HMIS data with point-in-
time data to enhance the information available on those experiencing 
homelessness. HHS has begun a project to get some of its homelessness 
programs to use HUD's HMIS system. For example, as discussed in more 
detail further on, in December 2009, HHS established an agreement with 
HUD requiring PATH providers to use HMIS. To address the issues faced 
by emergency shelters in quickly collecting and entering data on 
individuals, some Continuums issue identification cards containing 
demographic information to clients during their initial intake into 
the shelter system. Clients can swipe the cards as they enter a 
facility, and HMIS automatically captures the data. 

HUD and Others Have Used Data to Estimate the Extent of Homelessness, 
but Estimates Have Limitations: 

Despite the limitations discussed above, HUD uses data from point-in- 
time counts to estimate the number of those experiencing homelessness 
on a single night in January.[Footnote 19] HUD reported that 
approximately 660,000 individuals and persons in families experienced 
sheltered and unsheltered homelessness on a single night in January 
2008. However, this estimate does not include people who do not meet 
the definition of homelessness for HUD's programs but do meet 
definitions of homelessness for other programs. For example, HUD's 
counts would not include families living with others as a result of 
economic hardship, who are considered homeless by Education. Figure 1 
shows the count of sheltered and unsheltered persons experiencing 
homelessness on a single night in January for the past 4 years. 

Figure 1: Point-in-Time Count of Homeless Individuals and Persons in 
Families, 2005-2008: 

[Refer to PDF for image: vertical bar graph] 

Year: 2005; 
Sheltered: 418,165; 
Unsheltered: 344,845; 
Total: 763,010. 

Year: 2006; 
Sheltered: 427,971; 
Unsheltered: 331,130; 
Total: 759,101. 

Year: 2007; 
Sheltered: 391,401; 
Unsheltered: 280,487; 
Total: 671,888. 

Year: 2008; 
Sheltered: 386,361; 
Unsheltered: 278,053; 
Total: 664,414. 

Source: HUD, Annual Homeless Assessment Report to Congress, 2009. 

[End of figure] 

HUD also samples a number of communities and uses their HMIS data to 
estimate those experiencing homelessness in shelters during the year. 
[Footnote 20] HUD estimated that in 2008, 1.18 to 2 million people met 
HUD's definition of homelessness and were sheltered at some time in 
the year.[Footnote 21] The estimate has a broad range because HUD uses 
a sample of 102 communities and not all of those communities can 
provide usable data. For those Continuums related to the communities 
that can provide data on at least half of the beds in their inventory, 
HUD assumes that the remaining beds would be occupied in similar ways 
to estimate shelter use for those Continuums that cannot provide such 
data. HUD officials noted that response rates have been steadily 
improving and the estimate's range has decreased. For example, in 
2008, 87 of the 102 communities in HUD's sample provided usable data 
and another 135 communities voluntarily submitted data; while in 2005, 
55 communities in a sample of 80 communities provided usable data and 
another 9 communities contributed data voluntarily. HUD estimates that 
individuals without children make up about two-thirds and families 
with children under 18 about one-third of the estimate. However, HMIS 
only captures individuals and families who are defined as homeless 
under the McKinney-Vento Individual definition. Additionally, as 
previously noted, concerns exist about HMIS's ability to accurately 
record family status. 

HUD, HHS, and Education also report on other populations experiencing 
homelessness. HUD estimated that over the course of 2008, 
unaccompanied youths accounted for 2 percent of the sheltered homeless 
population, or approximately 22,000 unaccompanied youth who were 
homeless and sheltered. According to HHS, over the course of fiscal 
year 2008, approximately 48,000 youths experienced homelessness and 
received services from HHS' Basic Center Program or Transitional 
Living Program, which have different eligibility criteria from HUD's 
programs. Some youths may be in shelters funded by both HHS and HUD, 
and therefore be counted in both HMIS and RHYMIS, while others might 
be in shelters funded only by HUD or only by HHS and only included in 
the corresponding database. 

As shown in figure 2, Education reported that more than 770,000 
homeless children received services in the 2007-2008 school year, but 
less than one quarter of these children--about 165,000--were living in 
shelters. HUD reported for that same year that approximately 150,000 
children aged 6 to 17 were in shelters. 

Figure 2: Homeless Students Served by the Education of Homeless 
Children and Youth Program, by Primary Nighttime Residence Type and 
School Year: 

[Refer to PDF for image: illustrated table] 

Shelters: 
2005-2006: 207,925; 
2006-2007: 161,640; 
2007-2008: 164,982. 

Doubled up: 
2005-2006: 484,463; 
2006-2007: 420,995; 
2007-2008: 502,082. 

Unsheltered: 
2005-2006: 29,913; 
2006-2007: 54,422; 
2007-2008: 50,445. 

Hotels/motels: 
2005-2006: 65,420; 
2006-2007: 51,117; 
2007-2008: 56,323. 

Unknown residence: 
2005-2006: 91,864; 
2006-2007: N/A; 
2007-2008: N/A. 

Total: 
2005-2006: 879,594; 
2006-2007: 688,174; 
2007-2008: 773,832. 

Source: Department of Education, Education for Homeless Children and 
Youth. 

Estimating the Size of the Population Living With Extended Family or 
Nonfamily Due to Economic Hardship: 

While it is difficult to estimate the exact number of persons living 
with an extended family or nonfamily member because of economic 
hardship, evidence from several sources suggests that a substantial 
number of people may have this housing status. Education reported that 
502,082 school-age children identified as homeless under the McKinney-
Vento Children and Youth definition of homelessness were living 
doubled up during the 2007-2008 school year. The National Alliance to 
End Homelessness has estimated that 2.4 to 3.8 million people could 
have been doubled up for economic reasons in 2005. We analyzed data 
from the American Community Survey, an annual survey that collects 
population and housing data, to estimate that from 920,000 to 2.2 
million people were experiencing severe to moderate economic hardship 
and living with an extended family or nonfamily member in 2008.[A] 
This accounts for approximately 0.4 to 0.9 percent of the U.S. 
population. Of this estimate, approximately half are youths aged 16 
through 22. We made several assumptions about what comprises severe or 
moderate economic hardship. Severe economic hardship was assumed to 
mean that households had housing costs of at least 50 percent of 
household income that was below 50 percent of the federal poverty line 
and moderate economic hardship was assumed to mean that the households 
had housing costs that were at least 30 percent of household income 
that was below the federal poverty line. We also made assumptions 
about what comprises extended family; we assumed that extended family 
households were those where some people in the household were not part 
of the head of household's immediate family, and we included spouse, 
live-in partners, children, grandparents, and grandchildren in our 
definition of immediate family members. Our results would likely 
produce a different outcome if we made other assumptions. In addition, 
we cannot determine from the available data whether the individuals 
that are living with extended family or nonfamily members and 
experiencing severe or moderate economic hardship would meet the 
McKinney-Vento Children and Youth definition of homelessness which 
requires that individuals be doubled up because of economic hardship. 
As a result, we cannot determine whether the people in our estimate 
would be eligible for benefits if the McKinney-Vento Individual 
definition of homelessness were expanded to include those doubled up 
because of economic hardship. [A] These estimates have margins of 
error of less than plus or minus 4 percentage points of the estimate 
at the 95-percent confidence interval. 

Data from Mainstream Programs Could Improve Understanding of 
Homelessness, but Programs Have Not Consistently Collected Information 
on Housing Status: 

Federally-funded mainstream programs, whose primary purpose is to 
provide a range of services and funds to low-income households, often 
provide these services and funds to those who are experiencing or have 
experienced homelessness or to those defined as being at risk of 
becoming homeless. Thus, while homelessness is not the primary focus 
of these programs, data collected by them could be useful for 
understanding the nature of homelessness. Further, several researchers 
and advocates with whom we spoke noted that they could better 
understand the dynamics of homelessness if these programs collected 
individual client-level data on homelessness and housing status as 
part of their routine data collection activities. However, these 
programs have not consistently collected data on homelessness and 
housing status. A few programs have collected individual data, some 
have collected aggregate data, and others collect no data on housing 
status at all. 

We identified several federally-funded mainstream programs that 
collect or are beginning to collect and report client-level data on 
persons experiencing homelessness to the federal agency overseeing the 
mainstream program. Public Housing Authorities (PHA) collect data on 
homelessness status of households at the time the PHA admits the 
household to a housing assistance program, which includes both Public 
Housing and Housing Choice Voucher programs; they report these data to 
HUD's Office of Public and Indian Housing. HHS's Substance Abuse 
Prevention and Treatment Block Grant program requires grantees to 
report participants' living arrangements at entry and exit. DOL's WIA 
Adult and Youth grantees also collect and report individual-level data 
on enrolled participants including whether the client is homeless. 
HHS's John H. Chafee Foster Care Independence Program has developed a 
survey that states must begin using by October 2010 to gather data for 
the National Youth in Transition Database--a data collection required 
by the Foster Care Independence Act of 1999.[Footnote 22] States are 
required to survey foster care youths at ages 17, 19, and 21 to 
collect data on the services provided to, and outcomes of, youths in 
the foster care system. The survey includes a question asking youths 
if they have experienced homelessness over the relevant time period; 
however, as previously noted, the social stigma attached to the word 
homeless often limits self-identification. States administer USDA's 
SNAP program, document if a person or family is homeless, and report a 
sample of data to USDA, which uses the data to assess the accuracy of 
eligibility decisions and benefit calculations. 

A number of other programs require that grantees report aggregate data 
to their funding agency on the number of persons experiencing 
homelessness that they served: 

* Head Start grantees report the number of homeless families served 
annually to HHS. 

* Health Center Program grantees collect limited data on the 
homelessness status of program participants and report the total 
number of participants known to be homeless to HHS. 

* Community Mental Health Services Block Grant grantees collect and 
submit data to HHS on persons served by the program, including 
"homeless or shelter." 

* The Ryan White HIV/AIDS Program collects and reports to HHS limited 
aggregate data on the living arrangements (permanent, homeless, 
transient, or transitionally housed) of clients served. 

HHS has numerous other mainstream programs that provide funds to 
states to provide services to certain low-income populations, 
including those experiencing homelessness, but data collection and 
reporting on homelessness or housing status varies by program and 
across states. Medicaid and TANF are the two largest programs, but 
states are not required to collect or submit information to HHS on the 
number of individuals or families experiencing homelessness that they 
served. States determine eligibility requirements and develop program 
applications for TANF and Medicaid. A recent HHS study that surveyed 
all the states found that all states collected minimum housing status 
data on their Medicaid and TANF applications, such as home address and 
if the applicant resides in public or subsidized housing, a long-term 
care facility, or a medical or rehabilitation facility.[Footnote 23] 
Twenty eight states collected indicators of homelessness--such as 
whether an individual resides in a shelter, stays in a domestic 
violence shelter, or has a permanent home--on their applications. 
Thirteen states collected information on risk factors often associated 
with homelessness--such as whether an individual lives with friends or 
relatives, or has an eviction notice--on their applications. However, 
these states did not collect this information using consistent 
definitions and used the data in limited ways. According to the HHS 
report, most states responding to HHS's survey said that they did not 
know whether they had procedures in place to improve the quality of 
the items collected and thus how complete their homelessness data 
were. Additionally, while data on homelessness indicators and risk 
factors resided in statewide databases in many states, the data were 
not routinely confirmed or verified, making it unclear how reliable 
the data might be for analysis of homelessness. Further, as previously 
discussed, homelessness status changes over time, and data collected 
at one point in time may not accurately capture these changes. 
Nonetheless, in Michigan, New York City, and Philadelphia, researchers 
and state officials have been able to use identifying data in 
mainstream databases to match data in HMIS, and have thus been able to 
identify patterns in mainstream service usage for homeless populations. 

Several other mainstream programs provide services for persons 
experiencing homelessness, but do not provide aggregate or individual- 
level data on homeless clients served. The Community Services Block 
Grant, Social Services Block Grant, Maternal and Child Health Block 
Grant, and the Children's Health Insurance Program all provide HHS 
with regular program reports. However, these reports do not include 
data on the number of clients experiencing homelessness or other 
housing status data. Although child welfare agencies often collect 
data on housing status and stability in the process of reviewing 
family reunification cases, this information is not reported to HHS. 
Community Development Block Grants often fund services that may 
benefit those experiencing homelessness, but grantees do not track the 
number of homeless served by the program. Additionally, local PHAs can 
give preferences to individuals and families experiencing 
homelessness; however, PHAs do not have to submit data on these 
preferences to HUD. HUD sampled Public Housing and Housing Choice 
Voucher Program to determine how many of them have a preference for 
those experiencing homelessness. The analysis showed that in 2009, 
approximately 27 percent of all PHAs had a homeless preference. 

Finally, agencies have not always consistently collected or analyzed 
data on housing stability or homelessness because these are not the 
primary purposes of their programs. In addition, data collection may 
be expensive, and agencies must weigh the costs and benefits of 
getting more detailed information. Collecting data on homelessness or 
housing status for programs such as TANF and Medicaid could be further 
complicated by the need to work with 50 different state offices to 
implement a new data collection effort. However, HHS recently reported 
that of the 28 states that do collect homelessness data, almost all of 
them indicated that it is not burdensome or costly to collect such 
data, and about half of the states that collect data said they would 
comply with requests to make some homelessness data available to HHS 
for research purposes. Yet even among the willing states, there were 
some concerns about resource constraints for responding to such 
requests and concerns about the reliability of the data. However, not 
having complete and accurate data limits the understanding of the 
nature of homelessness--a better understanding of which could be used 
to inform programs and policies designed to improve housing stability 
and thus reduce homelessness. 

Definitional Differences and Data and Methodological Issues in 
Research Studies Hinder Development of Comprehensive Understanding of 
Factors Associated with Homelessness: 

Definitional Differences and Measurement Issues in Research Studies 
Make Analysis of Factors Associated with Homelessness Difficult: 

The 45 research studies analyzing factors associated with homelessness 
that we reviewed used different definitions or measurements of 
homelessness, although many of the studies used definitions or 
measures that were more closely affiliated with the McKinney-Vento 
Individual definition than with the broader McKinney-Vento Children 
and Youth definition (see appendix II for a list of the 45 studies). 
[Footnote 24] As a result, study findings are difficult to compile or 
compare. In the absence of a consistent definition and measurement, 
"homelessness" can mean or designate many conditions. For example, 
homelessness can refer to long-term homelessness, short stays in 
shelters, living in nontraditional housing, or living with relatives, 
friends, or acquaintances. These definitional differences especially 
limited research on some specific populations, such as "runaway or 
homeless" youths. 

The research we reviewed also varied in how it defined and measured 
the factors that may be associated with the likelihood of experiencing 
homelessness. For instance, studies that examined families and youths 
used different definitions or, in some cases, failed to clearly define 
what they meant by families and youth. Several studies measured 
variables such as marital status, social or family support, or 
domestic violence differently. For example, in assessing relationships 
between family structure and homelessness, one study examined whether 
a father of a child was cohabitating with a woman, while another study 
looked at whether the individual was presently married, although it is 
possible the two categories overlapped.[Footnote 25] Studies also used 
various age categories to define youths, including under 17, from 14 
to 23, or from 12 to 22. In addition, some studies did not consider 
factors that figured prominently in other studies, such as the 
economic conditions of the surrounding area or how childhood 
experiences influenced later episodes of homelessness. 

Data Limitations and Other Methodological Issues Limit What Overall 
Body of Research Can Say about the Factors Associated with 
Homelessness: 

To contribute to a broad-based and reliable understanding of what 
factors are associated with the likelihood of experiencing 
homelessness, studies we reviewed and experts with whom we spoke noted 
research would need to use data that accurately reflect the population 
studied, track the same individuals or families over time, and 
consider a broad population over diverse locations. Further, such 
studies would need to consider a range of both structural factors, 
such as area poverty level, and individual factors, such as a person's 
age. However, the majority of the studies we reviewed did not meet 
these criteria. As a result, the body of literature we reviewed cannot 
be used to predict with accuracy who among those at risk of 
homelessness would likely experience it.[Footnote 26] 

Studies we reviewed used samples from several types of data, such as 
providers' administrative databases or surveys, but were not always 
able to ensure that data accurately reflected the population they 
studied. Approximately half of the studies used information from 
administrative records or other service-oriented data, such as 
standardized self-assessments. The remaining studies used information 
collected in interviews, surveys, or questionnaires. Studies using 
administrative data may be especially vulnerable to biased sampling or 
undercounting of street homeless populations because of the myriad 
issues described previously, such as collecting data only on those 
receiving services. Some researchers noted that data from secondary 
sources such as administrative data may be less accurate than data 
collected by research staff and targeted for research purposes. 
However, survey data collected for research purposes also are subject 
to undercounting and biased sampling, because populations experiencing 
homelessness are difficult to reach. 

Because people move in and out of homelessness and experience it for 
different periods, studies we reviewed and experts with whom we spoke 
noted that data would need to be collected on the same individuals or 
families over time to more clearly identify which factors could lead 
to an episode of homelessness or help determine homelessness 
experiences over longer periods. Like HUD's point-in-time homeless 
counts, these studies more likely captured those individuals or 
families who had been homeless for long periods as opposed to those 
who experienced it episodically or for short periods, and thus do not 
give a clear understanding of factors associated with homelessness. 
These studies also could not determine whether factors associated with 
being homeless at a point in time caused homelessness. For example, 
one study found an association between poor physical health and 
homelessness, but could not say whether poor physical health 
contributed to experiencing homelessness or whether homelessness 
contributed to or worsened physical health. Nineteen studies in our 
review used data that did follow individuals or families over time. 
However, several of these used administrative data that suffered from 
the shortcomings described previously, followed individuals or 
families for relatively short periods, or considered populations in 
narrow geographic locations. A few studies also used national 
databases such as the Fragile Families and Child Wellbeing Study and 
one used the 1997 National Longitudinal Survey of Youth that annually 
tracks a sample of youth and their parents over time.[Footnote 27] 

In addition, most of the studies we reviewed defined their target 
populations--or the group of people to whom findings can be 
generalized--narrowly, making it difficult to generalize results to 
broader populations or to compare or compile them. Much of the 
research we reviewed focused on small subsets of the population 
experiencing homelessness in smaller geographic regions, such as those 
with mental illness or substance abuse problems in a single shelter or 
city. For example, one study published in a journal on Community 
Mental Health focused on African Americans admitted to a state 
psychiatric hospital in New York, and another study published in a 
journal on youth and adolescence looked at youths aged 14 to 21 years 
who needed the services of a homeless drop-in center.[Footnote 28] In 
part, the target groups studied reflected the wide variety of 
disciplines--psychology, public policy, public health, and economics--
of those conducting the studies. 

Although researchers have argued that it is necessary to consider 
structural or macro-level factors (such as employment rate, 
surrounding poverty level, and availability of affordable housing) as 
well as individual-level factors to arrive at a full understanding of 
which factors are associated with the likelihood of experiencing 
homelessness, only about one-third of the studies we reviewed 
considered these factors. Structural factors help to explain the 
prevalence of homelessness across a wider setting, while individual- 
level factors may explain the immediate circumstances surrounding an 
episode of homelessness. In addition, over three quarters of the 
service providers, researchers, advocates, and government agency 
officials we interviewed identified a structural factor--the lack of 
affordable housing--as a major barrier to serving those experiencing 
homelessness. However, most of the studies did not look at structural 
factors and focused on individual-level factors such as demographic 
characteristics, individual income, the presence of a mental illness, 
or substance abuse. Because the majority of the studies that we 
reviewed examined populations in one or a few cities, it was not 
possible for them to examine the role played by structural factors, 
such as unemployment rates and surrounding poverty levels, in a wider 
context. 

Despite Limitations in the Body of Literature, Studies Identified 
Factors That May Be Associated with Homelessness: 

Although limitations in the studies we reviewed posed challenges for 
drawing comparisons and often focused on narrow populations in smaller 
areas, we identified two that tracked homeless families over time and 
considered structural and individual-level factors across wide 
geographical areas. 

* One study that defined homelessness as living in a shelter, on the 
street, or in an abandoned building or automobile, but also considered 
the population that was doubled up, examined factors associated with 
individual and family homelessness using nationwide data from the 
Fragile Families and Child Wellbeing database, which was collected 
over several years.[Footnote 29] The study analyzed data on mothers 
when their children were one and three years old. One hundred and 
twenty-eight mothers reported experiencing homelessness at the one- 
year birthday, while 97 reported being homeless when their child 
turned three. A larger number of mothers reported being doubled up--
343 at their child's one-year birthday and 223 when their child turned 
three. The study found that the availability of affordable housing--a 
structural factor--reduced the odds of families experiencing 
homelessness and doubling up. A number of individual-level factors 
were associated with experiencing homelessness or doubling up. 
Specifically, access to small loans and childcare, having a strong 
family and friend support network, and living longer in a given 
neighborhood were associated with lowered odds of experiencing 
homelessness. Additionally, receiving public assistance reduced the 
likelihood that someone would live doubled up. 

* Another study considered families homeless if they were living on 
the street, in temporary housing, or in a group home, or had spent at 
least one night in a shelter or other place not meant for regular 
housing in the past 12 months.[Footnote 30] This study, which used the 
Fragile Families and Child Wellbeing database found that families with 
higher incomes who received housing assistance had a reduced 
likelihood of experiencing homelessness. Physical and mental health 
problems, reports of domestic violence, and substance abuse issues 
appeared to place them at greater risk for homelessness. Receipt of 
TANF and poorer surrounding economic conditions--a structural factor--
also were positively related to the likelihood of experiencing 
homelessness but, according to the authors, likely were proxies for 
individual need and lack of income rather than directly associated 
with homelessness. 

Two other studies looked at the association of structural factors and 
rates of homelessness across geographical areas over time, but did not 
track specific individuals or families: 

* One nationwide study that tracked homelessness rates over time 
primarily examined how structural factors affected rates of 
homelessness.[Footnote 31] The study found that relatively small 
changes in housing market conditions could have substantial effects 
upon rates of homelessness or the numbers of persons in shelters. 
Their results imply that relatively small increases in housing vacancy 
rates, combined with small decreases in market rents, could 
substantially reduce homelessness. 

* Another study that focused on the impact of structural factors on 
homelessness in 52 metropolitan areas found that poverty levels 
strongly related to the number of persons experiencing homelessness in 
an area.[Footnote 32] No other structural factors--such as 
unemployment rates, the number of government benefit recipients, or 
availability of affordable housing in the area--were found to be 
statistically significant predictors of homelessness. 

Together, the four studies underlined the importance of structural 
factors and identified some individual factors associated with 
homelessness; however, they did not address some issues of importance. 
None addressed the extent to which childhood experiences were 
associated with adult homelessness, and only one examined those living 
in doubled up situations. 

We reviewed 11 other studies that examined how childhood experiences 
were associated with experiencing homelessness in adulthood; however, 
these studies generally relied on people's recollections.[Footnote 33] 
While the studies used varying methodologies and definitions of 
homelessness and other factors, most highlighted the influence of 
early childhood experiences on the likelihood of later experiencing 
homelessness. Results varied by study, but several studies found that 
factors such as running away from home, being in foster care, having a 
dysfunctional family, or being sexually molested as a child increased 
the odds an adult would experience homelessness. Similarly in 1996, 
the National Survey of Homeless Assistance Providers and Clients found 
that homeless adults reported many significant adverse childhood 
experiences. That survey did not compare those experiencing 
homelessness with those that were not. However, the findings from the 
studies we reviewed that did compare the two groups generally were 
consistent with the survey's findings. Conversely, another study found 
that a range of childhood experiences (including residential 
stability: adequacy of income; dependence on public assistance; family 
violence; and parental criminality, mental illness, or substance 
abuse) were not significantly associated with adult homelessness. 
[Footnote 34] 

Recognizing that the relationships between living doubled up or in 
shelters or on the street are important to understanding homelessness, 
we identified a few studies that analyzed whether doubling up could 
predict future time spent in a shelter or on the street, or that 
measured differences at a point in time between those living doubled 
up and those living in shelters or on the street. However, the results 
of the studies were inconclusive. Of the two that examined how 
doubling up affects later homelessness in a shelter or on the street, 
one found that it was significant and the other found it was not 
significant.[Footnote 35] Of the four studies that compared persons on 
the street or in a shelter with those doubled up, two found few 
differences in demographic characteristics or backgrounds.[Footnote 
36] A third found some differences between the two groups.[Footnote 
37] For example, receiving public assistance lowered the chance of 
doubling up but was not significantly associated with homelessness. 
The fourth study found significant differences between doubled up and 
homeless mothers. Doubled up mothers were more likely to be younger 
and working and to have high school degrees, fewer children, and more 
relatives who could help with finances, housing, and child care. 
[Footnote 38] 

Definitional Issues Have Posed Challenges for Service Providers and 
Make Collaborating at Local and Federal Levels More Difficult: 

Definitional Differences Limited Providers' Ability to Serve Certain 
Populations Effectively: 

Among the majority of the advocates, government officials, service 
providers, and researchers we interviewed that identified differences 
in definitions of homelessness as an important barrier to providing 
services, several noted that families and youths living in some 
precarious situations were not eligible for federal assistance under a 
narrow definition of homelessness. Some said that families and youths 
who were doubled up or living in hotels because of economic hardship 
often had similar or greater needs for services than those who met 
narrower definitions, but were being excluded from receiving 
government-funded services. For example, those working in educational 
programs that have broader federal definitions of homelessness noted 
that those who do not meet the narrow definition have difficulty 
accessing housing services. One of the school liaisons we visited 
described visiting a house with a caved-in floor and no front door. 
This family met the criteria of substandard housing under the McKinney-
Vento Children and Youth definition of homelessness, but it is unclear 
whether the house would have been considered abandoned or condemned, 
and if the family would have qualified as homeless under the narrower 
individual definition prior to the HEARTH Act. According to a research 
study presented at the HUD-HHS homelessness research symposium in 
2007, a formal condemnation process for substandard properties does 
not typically exist in rural areas, and, as a result, properties that 
would meet the HUD definition of abandoned because they have been 
condemned in urban areas may not meet that definition in rural 
areas.[Footnote 39] HHS provides grants for Head Start programs to 
collaborate with others in the community to provide services for 
children and their families; however, officials noted that in the 2009 
program year, less than half of the families in Head Start who 
experienced homelessness acquired housing. HHS has attributed this to 
a lack of affordable housing and long waiting lists for housing 
assistance. However, officials for at least one service provider said 
that the waiting list for housing assistance in their city was much 
longer for those that do not meet the narrow definition of 
homelessness. 

Many of those involved in homeless programs with whom we spoke were 
particularly concerned about the exclusion of families and youths from 
programs that addressed the needs of chronically homeless individuals--
those unaccompanied individuals who have disabilities and have been 
continuously homeless for a year or homeless four times in the last 3 
years. Before the passage of the HEARTH Act, families that otherwise 
met the criteria for chronic homelessness programs were not able to 
participate because chronic homelessness was defined to include only 
unaccompanied individuals.[Footnote 40] People in all of the 
categories we interviewed noted that the emphasis on funding programs 
for chronic homelessness has meant that families have been 
underserved. A youth service provider further noted that youths 
effectively were excluded from programs for those experiencing chronic 
homelessness because youths generally did not live in shelters or keep 
records of where they had been living or for how long. 

Those that cited differences in definitions as a barrier said that 
families and youths with severe shelter needs had to be on the street 
or in shelters to access some federally-funded homeless assistance, 
but shelters were not always available or appropriate for them. 
Researchers we interviewed noted that families have to obey a number 
of rules to stay in a shelter and families with the greatest 
challenges might be less able to follow those rules. Additionally, 
some facilities do not provide shelter for males above a certain age, 
so that couples or families with male teenage children may not be able 
to find shelter together. Similarly for youths, a researcher and a 
service provider suggested that adult shelters were not appropriate 
for unaccompanied youths or young adults, and shelters specifically 
for them were very limited. 

Some of the people we interviewed also noted that some narrow 
definitions of homelessness limited services that could be provided to 
individuals experiencing homelessness. For example, getting one 
service sometimes precluded individuals from getting another service 
for which they would otherwise have qualified. Officials at DOL told 
us that if veterans obtain housing vouchers through HUD-VASH, they no 
longer meet the narrow statutory definition of homelessness under 
which they would be eligible for job training funded by the Homeless 
Veterans Reintegration Program (HVRP). However, if veterans first 
apply for HVRP and then for vouchers, they can qualify for both 
programs. Similarly, those in transitional housing programs cannot be 
considered eligible for programs serving those experiencing chronic 
homelessness even if they meet the other requirements, such as being 
homeless for a year and having a disability. In addition, although HUD 
has recognized in its documents that helping people make successful 
transitions to the community as they are released from foster care, 
jails, prisons, and health care, mental health, or substance abuse 
treatment facilities requires systems to work together to ensure 
continuity of care and linkages to appropriate housing and community 
treatment and supports, the definitions of homelessness may hinder 
these transitions. In August 2009, one advocate noted that HUD's 
definition of homelessness includes those that spend 30 days or less 
in prison if they had been homeless prior to entering prison, but 
those spending more than 30 days cannot be considered homeless until 
the week before their release.[Footnote 41] The advocate said that 
this limits the incentive for prison staff to work with homeless 
service providers to allow for a smooth transition from prison to 
housing and that if an individual leaving prison spends time on the 
street or in an emergency shelter, the likelihood of recidivism 
increases. 

Some of those arguing for a broader definition also have said that the 
definition of homelessness should not depend on available funding. 
Officials at one large service provider said that broadening the 
definition would not necessarily spread a fixed amount of resources 
across a larger group. Instead, targeting resources to specialized 
populations more effectively and concentrating on earlier intervention 
and prevention could lower the cost of serving individual clients. 
However, they also noted that this might require a better 
understanding of the needs of particular subgroups experiencing 
homelessness. 

Some local officials, homeless service providers, and researchers 
noted that choosing between a narrow or a broad definition of 
homelessness was less important than agreeing on a single definition, 
because multiple definitions made it more difficult or costly to 
provide services and created confusion that sometimes led to services 
not being provided to those legally eligible for them. Many 
researchers, government officials, and advocates with whom we spoke 
noted the importance of combining services and housing to meet the 
needs of those experiencing homelessness, and some of these noted that 
this was more difficult and costly when programs defined homelessness 
differently. They also noted that obtaining funding for services from 
sources other than HUD has become more necessary because the 
proportion of HUD funding for services has declined. Officials at HUD 
noted that this was a result of HUD having provided incentives to 
communities to increase the ratio of housing activities to supportive 
service activities in their funding applications to encourage the 
development of more housing resources for individuals and families 
experiencing homelessness. 

Not only do some targeted programs that provide services use different 
definitions of homeless, but some state and local grantees receiving 
federal funds under mainstream programs that can be used to provide 
certain services for those experiencing homelessness (such as TANF) 
develop their own local definitions of homelessness. Officials at a 
lead Continuum agency said that having these different definitions 
makes putting together funding for permanent supportive housing--the 
best solution for ending chronic homelessness--especially difficult. 
Officials at two entities that provide service to and advocate for 
those experiencing homelessness noted that, given the multiple 
definitions, scarce resources that could have been used to provide 
services instead went to eligibility verification. Furthermore, many 
of those involved in activities related to homelessness said that 
having multiple definitions created confusion, and government 
officials overseeing programs that use a broader definition and a 
service provider in one of these programs noted that this confusion 
could lead to services not being provided to those that are eligible 
for them. A school liaison and a youth service provider said that 
school administrative personnel often apply a narrower definition of 
homelessness than McKinney-Vento Children and Youth and thus may deny 
students access to services to which they are entitled. Additionally, 
Education has cited a state education agency for the failure of local 
education agencies' to identify, report, and serve eligible homeless 
children and youths including youths in doubled-up situations that 
meet the broader definition of homelessness. Similarly, officials at 
HHS acknowledged that Head Start programs across the country sometimes 
were not using the appropriate definition of homelessness to identify 
children who qualified for those services. As a result, some homeless 
families would not be receiving Head Start services. 

However, some government officials, researchers, advocates, and 
service providers thought that having multiple or narrow definitions 
of homelessness had certain benefits. Some HHS officials in programs 
that address homelessness and others noted that having multiple 
definitions of homelessness allowed programs to tailor services and 
prioritize them for specific populations. HUD officials and some 
researchers and advocates said that having a narrow definition for 
homeless programs that provide shelter for specific populations and 
broader definitions for programs such as those designed to serve the 
educational needs of children and youths allowed programs to meet 
their goals best. HUD officials noted that having a broader definition 
for certain education programs is appropriate because those that meet 
the definition are entitled to the service, and the program does not 
provide housing. Alternatively, it is appropriate for programs such as 
HUD's to have a narrower definition because its services are not 
entitlements and must target those most in need, such as those that 
are chronically homeless. HUD, HHS, VA, and DOL began redirecting 
resources to this narrowly defined group in 2003, and according to HUD 
point-in-time data, chronic homelessness fell by approximately 27 
percent from the January 2005 count to the January 2008 count. HUD, 
HHS, and VA focused on this group, in part, because a research study 
had shown that they used an inordinate amount of shelter 
resources.[Footnote 42] One researcher noted that having a precise 
definition was essential to ensure that the same kinds of people are 
being counted as homeless in different locations, which would be 
important for measuring program outcomes. Supporters of a narrow 
definition also said that if the definition were broadened, limited 
resources might go to those who were easier to serve or had fewer 
needs, specifically to those families with young children who were 
doubled up rather than to those identified as chronically homeless. 
Finally, some advocates for those experiencing homelessness and 
government officials overseeing programs targeted at those 
experiencing homelessness noted that if the definition of homelessness 
were broadened for some programs without an increase in resources, 
many of those that would become eligible for services would not get 
them. 

In the HEARTH Act, Congress provided a broader definition of 
homelessness for those programs that had been serving individuals and 
families and using the McKinney-Vento Individual definition; however, 
it is still not as broad as the McKinney-Vento Children and Youth 
definition, so different definitions will still exist when the HEARTH 
Act is implemented. In addition, the HEARTH Act mandated that the 
Interagency Council convene experts for a one-time meeting to discuss 
the need for a single federal definition of homelessness within 6 
months of the issuance of this report.[Footnote 43] However, having 
one definition of homelessness would not necessarily mean that 
everyone who met that definition would be eligible for all homeless 
assistance programs or that those not defined as homeless would be 
ineligible. Some of the people we interviewed suggested alternatives--
one based on a narrow definition of homelessness and others based on a 
broader definition. For example, one local official suggested defining 
homelessness using the narrow McKinney-Vento Individual definition and 
defining another category called "temporarily housed" that would 
include those who are doubled up or in hotels. While some programs 
might only be open to those experiencing homelessness, others such as 
the Education of Homeless Children and Youth program could be open to 
both groups. Alternatively, one researcher directed us to a 
classification scheme developed by the European Federation of National 
Associations Working with the Homeless. Under that classification 
scheme, homelessness was defined broadly as not having a suitable home 
or one to which a person was legally entitled, but then a typology was 
created that defined subcategories of living situations under headings 
such as "roofless" or "inadequate" that could be addressed by various 
policies. Officials at a large service provider we interviewed made 
similar distinctions saying that it is best to think of people as 
experiencing functional homelessness--that is, living in situations 
that could not be equated to having a home--rather than to think of 
them as literally homeless or doubled up. However, these officials 
said that subcategories of need would have to be developed based on a 
better understanding of homelessness, because all persons experiencing 
homelessness should not be eligible for the same services. In 2007, 
HHS convened a symposium to begin discussing the development of a 
typology of homeless families, and in May 2010, they convened about 75 
federal and nonfederal participants to discuss issues related to 
children experiencing homelessness. 

Those Experiencing Homelessness Have Faced Other Barriers, Especially 
the Lack of Affordable Housing and Difficulty Accessing Mainstream 
Programs: 

The lack of affordable housing (whether housing was not available or 
people's incomes were not high enough to pay for existing housing) was 
the only barrier to serving those experiencing homelessness cited more 
frequently by researchers, advocates, service providers, and 
government officials we interviewed than definitional differences. 
Some researchers have shown that more housing vouchers could help 
eradicate homelessness, but a research study also has shown that 
generally federal housing subsidies are not targeted to those likely 
to experience homelessness. Those with certain criminal records or 
substance abuse histories may not be eligible for federal housing 
assistance, and these factors sometimes are associated with 
homelessness. Although certain federal programs target vouchers to 
those who are most difficult to house, local service providers may 
still refuse to serve those who have been incarcerated or have 
substance abuse problems. For example, while the HUD-VASH program is 
to be available to many of these subpopulations, HUD officials and 
others told us that local service providers still refuse to serve 
them. In addition, while HUD estimates that 27 percent of PHAs have 
preferences for those experiencing homelessness, many of them restrict 
these programs to those who may be easier to serve. 

Service providers, advocates, researchers, and government officials 
that we interviewed also cited eligibility criteria for mainstream 
programs as a main barrier to serving those experiencing homelessness. 
In 2000, we reported on barriers those experiencing homelessness faced 
in accessing mainstream programs, and this is a continuing issue. 
[Footnote 44] To obtain benefits, applicants need identification and 
other documents, which those experiencing homelessness often do not 
have. Without documentation, they sometimes cannot enter federal and 
state buildings where they would need to go to get documentation or 
obtain benefits. Those that cited access as a barrier particularly 
noted difficulties with SSI/SSDI programs. Service providers and 
government officials noted that those experiencing homelessness may 
not receive notices about hearing dates or other program requirements 
because they lack a fixed address. At least one researcher told us 
that an initiative, SSI/SSDI Outreach, Access and Recovery (SOAR), has 
improved performance. The initiative's Web site says that those 
experiencing homelessness normally have a 10-15 percent chance of 
receiving benefits from an initial application, but that SOAR has 
increased success to 70 percent in areas it serves. However, one local 
agency in an area served by SOAR told us in January 2010 that most 
applicants were rejected initially. Some of those we interviewed also 
noted that Medicaid applicants have some similar problems. For 
example, one advocate noted that it is difficult for those 
experiencing homelessness to get through the application process and, 
when necessary, prove disability; however, because Medicaid is a state-
run program, these problems are worse in some states than in others. 
Another provider noted that Medicaid requires that information be 
periodically updated, and those experiencing homelessness may not 
receive notices of this. As a result, they may lose their benefits and 
be required to travel a long distance to get them reinstated. Finally, 
service providers said that PHAs often restrict federal housing 
assistance to those without substance abuse issues or certain criminal 
records and that programs generally have long waiting lists. 

Definitional Issues Make Collaboration at Local and National Levels 
More Difficult: 

Because homelessness is a multifaceted issue and a variety of programs 
across a number of departments and agencies have been designed to 
address it, collaborative activities are essential to reducing 
homelessness in a cost-effective manner. In prior work, we have 
determined that certain key activities, such as setting common goals, 
communicating frequently, and developing compatible standards, 
policies, procedures, and data systems, characterize effective 
interagency collaboration.[Footnote 45] In addition, we found that 
trust is an important factor for achieving effective collaboration. 

Efforts to address homelessness often have stressed the need for 
local, communitywide collaboration. For instance, entities applying to 
HUD for Homeless Assistance Grants have to come together as a 
Continuum to file applications. Other agencies or individuals, such as 
the school systems' homeless liaisons, also are required to coordinate 
activities in the community. In addition, from 2002 to 2009, 
Interagency Council staff encouraged government officials, private 
industry, and service providers to develop 10-year plans to end 
homelessness or chronic homelessness and provided tools to communities 
to assist with the development of these plans. Many communities have 
developed these plans, but whether plans have been implemented or have 
been achieving their goals is unclear. The Interagency Council reports 
that 332 of these plans have been drafted. All of the locations we 
visited had drafted plans at the state or local level, however, in two 
of the four sites--California and South Carolina--plans that had been 
drafted had not been adopted by appropriate local or state government 
entities and thus had not been implemented. 

Some of the people with whom we spoke said that differences in 
definitions of homelessness limited their ability to collaborate 
effectively or strategically across communities. Local officials or 
researchers in three of the four locations we visited noted that 
certain elements of collaboration were difficult to achieve with 
different definitions of homelessness. In one location we visited, 
local agency officials who had extensive experience with a broad range 
of homelessness programs and issues noted that multiple definitions 
impeded those involved in homelessness activities from defining or 
measuring a common problem and were a major obstacle to developing 
measures to assess progress in solving the problem. Further, they 
noted that the trust of the local community in officials' ability to 
understand the problem of homelessness was eroded when recent point-in-
time counts showed that numbers of families experiencing homelessness 
under one definition declined while the number of families receiving 
homeless services in other programs that defined homelessness more 
broadly increased. In two other locations, local government officials 
and a researcher involved in evaluating local programs said that 
having multiple definitions of homelessness impeded their ability to 
plan systematically or strategically for housing needs or efforts to 
end homelessness at the community level. 

Congress also recognized the importance of federal interagency 
collaboration when it authorized the Interagency Council in the 
original McKinney-Vento Act and reauthorized it in the HEARTH Act. 
Some of the people we interviewed further noted that collaboration 
among federal programs was essential because addressing homelessness 
required that those in need receive a holistic package of services 
that might encompass the expertise and programs of a number of 
agencies. They also said that collaboration was necessary to prevent 
people from falling through gaps created by certain events, such as 
entering or leaving hospitals or prisons, aging out of foster care or 
youth programs, or otherwise experiencing changes in family 
composition. Further, they noted that, with HUD's emphasizing housing 
rather than services in its funding priorities, the need for effective 
collaboration was greater now than in the past. Finally, officials at 
HUD, HHS, and Education noted that at a time of budget austerity 
collaboration among agencies was an effective way to leverage scarce 
resources. 

While we noted in 1999 and again in 2002 that homeless programs could 
benefit from greater interagency coordination, many of the government 
officials, researchers, advocates, and service providers we 
interviewed who were knowledgeable about multiple federal agencies 
said that collaboration among federal programs and agencies had been 
limited or did not exist at all.[Footnote 46] Generally, those we 
interviewed in our current work said that, from 2002-2009, the 
Interagency Council had focused on that part of its mission that 
required it to foster local collaboration rather than on that part 
that required it to foster collaboration among federal agencies. In 
addition, some of those we interviewed said that federal program staff 
had focused largely on their own requirements and funding streams 
rather than on collaborative approaches to addressing homelessness. 

In 1994, the Interagency Council issued an interagency plan to address 
homelessness that called for federal agencies to streamline and 
consolidate programs, when appropriate, and introduced the concept of 
a Continuum of Care, but did not include any longer-term mechanism to 
promote interagency collaboration, such as joint funding of programs. 
Following issuance of this plan, the Interagency Council did not again 
receive funding until 2001, although it did undertake some joint 
activities including coordinating and funding a survey of service 
providers and persons experiencing homelessness. In 2002, an executive 
director was appointed and, according to some of those involved with 
the Interagency Council, the council turned its attention largely to 
helping communities draw up 10-year plans to end chronic homelessness. 
In the HEARTH Act, Congress called on the Interagency Council to 
develop a strategic plan to end homelessness that would be updated 
annually, and in November 2009, a new executive director took office. 
In preparation for the strategic plan and in response to new staffing 
and funding at the Interagency Council and elsewhere, agencies and the 
Council appear more focused on interagency coordination. The 
Interagency Council issued its strategic plan on June 22, 2010. 
[Footnote 47] The plan says that it is designed to neither embrace nor 
negate any definition of homelessness being used by a program. 

Federal agencies have also not collaborated effectively outside the 
Interagency Council. Those we interviewed noted that agencies have 
focused on their own funding streams and have not coordinated dates 
for applying for grants that could be combined to provide housing and 
services for those experiencing homelessness. Service providers must 
apply for grants at different times, and grants run for different 
periods and have different probabilities of being continued. A 
provider might receive funding to build permanent housing but might 
not receive funding needed for certain support services, or vice 
versa. One group knowledgeable about an array of housing programs said 
that recently an HHS grant tried to link its funding to HUD's, but a 
lack of full collaboration between the agencies created confusion and 
discouraged some service providers from applying for the HHS grant. 
The HHS grant required that applicants have an executed grant from HUD 
when they applied for the HHS grant. However, HHS applications were 
due before HUD had executed any of its grants. HHS officials then 
relaxed their grant criteria, saying that they would evaluate the lack 
of an executed grant contract with HUD on a case-by-case basis. HUD 
officials said that the grant criteria were relaxed to include 
recognition of HUD's conditional grant award letters. Two groups with 
whom we spoke also noted that funding from multiple agencies often 
focused on demonstration projects and that grant processes for these 
also were not well coordinated and funding ended abruptly. Officials 
at HUD noted that lack of coordination on grants across agencies is 
likely the result of the statutes that authorize programs and agency 
regulations that implement them. 

Some of the service providers, advocates, and government officials we 
interviewed cited specific examples of successful programmatic 
collaboration, such as the HUD-VASH program, and federal agency 
officials directed us to a number of initiatives that illustrate a 
greater emphasis on interagency collaboration.[Footnote 48] HUD 
officials noted that they have been partnering with HHS and VA to 
improve and align their data collection and reporting requirements for 
federally-funded programs addressing homelessness. For example, HUD 
and HHS announced in December 2009 plans to move toward requiring that 
HHS's PATH program use HMIS for data collection and reporting for 
street outreach programs. They noted that the agencies had agreed to 
align reporting requirements by establishing common outputs and 
performance outcomes. The plan called for HHS to begin providing 
technical assistance and training activities for PATH programs on 
individual-level data collection and reporting and alignment with HMIS 
in 2010, and to seek approval for a revised annual report to include 
HMIS data in 2011. In February 2010, officials from HUD, HHS, and 
Education--key agencies for addressing homelessness for nonveterans-- 
outlined proposals on homelessness included in the proposed FY 2011 
budget. These included a demonstration program that combines 4,000 HUD 
housing vouchers with HHS supportive services and another program that 
calls for HUD, HHS, and Education to be more fully engaged in 
stabilizing families. The latter proposal calls for HUD to provide 
6,000 housing vouchers on a competitive basis. 

We also found that federal agency staff did not effectively 
collaborate within their agencies. For example, in January 2010, staff 
at one of HUD's field offices told us that while collaboration between 
those involved in the Homeless programs and those involved in Public 
Housing programs would be beneficial, any coordination between these 
two HUD programs was "haphazard." In February 2010, the Assistant 
Secretaries for the Offices of Public and Indian Housing and Community 
Planning and Development, which includes homeless programs, reported 
that they are meeting weekly and looking for ways to better coordinate 
programs. In another example, staff at HHS who developed the National 
Youth in Transition Database, which includes looking at experiences 
with homelessness, had not consulted with staff in the Family and 
Youth Services Bureau, who administer the Runaway and Homeless Youth 
Programs and generally were recognized as having some expertise on 
youths experiencing homelessness. 

Finally, we observed that while coordination has been limited, it was 
more likely to occur between those parts of agencies that were using a 
common vocabulary. For example, state McKinney-Vento education 
coordinators and local education liaisons are required to coordinate 
with housing officials and providers in a number of ways; however, the 
McKinney-Vento Homeless Education Program coordinator in one of the 
states we visited said that while she has coordinated locally with 
staff from Head Start, an HHS program that also uses the McKinney-
Vento Children and Youth definition of homelessness, she has found it 
very hard to coordinate with local HUD staff that use a different 
definition of homelessness, because they did not see how the education 
activities relate to their programs. In addition, those agencies that 
have agreed on a definition of chronic homelessness--HUD, HHS, DOL, 
and VA--have engaged in some coordinated efforts to address the needs 
of those that met the definition. 

Conclusions: 

For many years, the federal government has attempted to determine the 
extent and nature of homelessness. As part of this effort, Education, 
HHS, and HUD have systems in place that require service providers 
involved in the homelessness programs they administer to collect data 
on those experiencing homelessness and report these data in various 
ways to the agencies. However, while the data currently being 
collected and reported can provide some useful information on those 
experiencing homelessness, because of difficulties in counting this 
transient population and changes in methodologies over time, they are 
not adequate for fully understanding the extent and nature of 
homelessness. In addition, the data do not track family composition 
well or contribute to an understanding of how family formation and 
dissolution relate to homelessness. Further, because of serious 
shortcomings and methodologies that change over time, the biennial 
point-in-time counts have not adequately tracked changes in 
homelessness over time. While these data systems have improved, it 
still is difficult for agencies to use them to understand the full 
extent and nature of homelessness, and addressing their shortcomings 
could be costly. For example, one shortcoming of HUD's point-in-time 
count is that it relies on volunteer enumerators who may lack 
experience with the population, but training and utilizing 
professionals would be very costly. 

In part because of data limitations, researchers have collected data 
on narrowly defined samples that may not be useful for understanding 
homelessness more generally or do not often consider structural 
factors, such as area poverty rates, which may be important in 
explaining the prevalence and causes of homelessness. In addition, 
because complete and accurate data that track individuals and families 
over time do not exist, researchers generally have not been able to 
explain why certain people experience homelessness and others do not, 
and why some are homeless for a single, short period and others have 
multiple episodes of homelessness or remain homeless for a long time. 

However, those who have experienced or might experience homelessness 
frequently come in contact with mainstream programs that are 
collecting data about the recipients of their services. While 
homelessness is not the primary focus of these programs, if they 
routinely collected more detailed and accurate data on housing status, 
agencies and service providers could better assess the needs of 
program recipients and could use these data to help improve the 
government's understanding of the extent and nature of homelessness. 
Researchers also could potentially use these data to better define the 
factors associated with becoming homeless or to better understand the 
path of homelessness over time. Collecting these data in existing or 
new systems might not be easy, and agencies would incur costs in 
developing questions and providing incentives for accurate data to be 
collected. Collecting such data may be easier for those programs that 
already collect some housing data on individuals, families, and youths 
who use the programs and report those data on an individual or 
aggregate basis to a federal agency, such as HHS's Substance Abuse 
Treatment and Prevention Block Grant program or Head Start. For those 
mainstream programs that do not currently report such data, collecting 
it may be a state or local responsibility, and the willingness of 
states to collect the data may vary across locations. For example, HHS 
has reported that about half of the states that do collect 
homelessness data do not consider it burdensome to do so through their 
TANF and Medicaid applications, and would be willing to provide data 
extracts to HHS for research purposes. States or localities and 
researchers could find these data useful even if they are not 
collected on a federal or national level. However, concerns exist 
about resource constraints and data reliability. Therefore, the 
benefits of collecting data on housing status for various programs 
would need to be weighed against the costs. 

Federal efforts to determine the extent and nature of homelessness and 
develop effective programs to address homelessness have been hindered 
by the lack of a common vocabulary. For programs to collect additional 
data on housing status or homelessness or make the best use of that 
data to better understand the nature of homelessness, agencies would 
need to agree on a common vocabulary and terminology for these data. 
Not only would this common vocabulary allow agencies to collect 
consistent data that agencies or researchers could compile to better 
understand the nature of homelessness, it also would allow agencies to 
communicate and collaborate more effectively. As identified in 2011 
budget proposals, Education, HHS, and HUD are the key agencies that 
would need to collaborate to address homelessness, but other agencies 
that also belong to the Interagency Council--a venue for federal 
collaborative efforts--such as DOL and DOJ might need to be involved 
as well. However, agency staff may find it difficult to communicate at 
a federal or local level when they have been using the same terms to 
mean different things. For example, agencies might want to avoid using 
the term homelessness itself because of its multiple meanings or the 
stigma attached to it. Instead, they might want to list a set of 
housing situations explicitly. The agencies could begin to consider 
this as part of the proceedings Congress has mandated that the 
Interagency Council convene after this report is issued. Once agencies 
have developed a common vocabulary, they might be able to develop a 
common understanding of how to target services to those who are most 
in need and for whom services will be most effective. In addition, 
with a common vocabulary, local communities could more easily develop 
cohesive plans to address the housing needs of their communities. 

Recommendations for Executive Action: 

To improve their understanding of homelessness and to help mitigate 
the barriers posed by having differences in definitions of 
homelessness and related terminology, we recommend that the 
Secretaries of Education, HHS, and HUD--working through the U. S. 
Interagency Council on Homelessness--take the following two actions: 

1. Develop joint federal guidance that establishes a common vocabulary 
for discussing homelessness and related terms. Such guidance may allow 
these and other agencies on the Interagency Council on Homelessness to 
collaborate more effectively to provide coordinated services to those 
experiencing homelessness. 

2. Determine whether the benefits of using this common vocabulary to 
develop and implement guidance for collecting consistent federal data 
on housing status for targeted homelessness programs, as well as 
mainstream programs that address the needs of low-income populations, 
would exceed the costs. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to the Departments of Education, 
Health and Human Services, Housing and Urban Development, Labor, and 
Justice and the Executive Director of the Interagency Council for 
their review and comment. We received comments from the Assistant 
Secretary of the Office of Elementary and Secondary Education at the 
Department of Education; the Assistant Secretary for Legislation at 
the Department of Health and Human Services; the Assistant Secretary 
of Community Planning and Development at the Department of Housing and 
Urban Development; and the Executive Director of the Interagency 
Council. These comments are reprinted in Appendixes III through VI of 
this report respectively. The Departments of Labor and Justice did not 
provide formal comments. 

Education, HUD, and the Executive Director of the Interagency Council 
explicitly agreed with our first recommendation that Education, HHS, 
and HUD--working through the Interagency Council--develop federal 
guidance that establishes a common vocabulary for discussing 
homelessness and related terms. HHS did not explicitly agree or 
disagree with this recommendation. Instead, HHS commented extensively 
on the advantages of having multiple definitions of homelessness. 
While we discuss the challenges posed by, and the advantages of, 
having multiple definitions of homelessness in this report, our report 
recommends a common vocabulary rather than either a single or multiple 
definitions. In their interagency strategic plan to prevent and end 
homelessness issued on June 22, 2010, the agencies acknowledge the 
need for a common vocabulary or language when they say that a common 
language is necessary for the interagency plan to be understandable 
and consistent and that this language does not negate or embrace the 
definitions used by different agencies. 

Education explicitly addressed our second recommendation that agencies 
consider the costs and benefits of using a common vocabulary to 
develop and implement guidance for collecting consistent federal data 
on housing status for targeted homelessness and mainstream programs in 
their written response. Education wrote that a discussion of such 
costs and benefits of using more of a common vocabulary, as it relates 
to data collection, should be an agenda item for the Interagency 
Council. The Executive Director of the Interagency Council also 
supported further exploration of how to accurately and consistently 
report housing status in mainstream programs. Although we recommend 
that the agencies work through the council to address this 
recommendation, decisions about individual program data collection 
will necessarily be made by the agency overseeing the program. 
Although HHS did not comment explicitly on our second recommendation, 
they did provide comments on data collection. They commented that GAO 
appears to assume that programs identify people who are homeless only 
to have a total count of the homeless population. We do not make that 
assumption. We recognize that programs collect data specifically for 
the program's use; however, data collected for programs also can 
contribute to a broader understanding of the extent and nature of 
homelessness. For example, while HMIS has certain shortcomings 
described in the report, service providers collect HMIS data in some 
cases to better manage their programs, and HUD also uses those data to 
attempt to understand the extent and nature of homelessness. HHS also 
noted that homelessness data systems are costly and complicated to 
develop and linking them presents challenges. We acknowledge that 
while collecting more consistent data on housing status for targeted 
and mainstream programs would have benefits, there would be 
implementation costs as well. 

Additionally, HHS, HUD, and the Executive Director of the Interagency 
Council raised other concerns about this report that did not relate 
directly to the two recommendations. HHS commented on the history of 
the National Youth in Transition Database, developed in response to 
the Chafee Foster Care Independence Act of 1999. HUD commented that 
the report did not present a complete view of HUD's data collection 
and reporting efforts and did not recognize the strides that have been 
made in this area, the value of the data currently being collected and 
reported, or that their Annual Homeless Assessment Report is the only 
national report to use longitudinal data. The Executive Director of 
the Interagency Council also wrote that the report did not adequately 
recognize what is possible today that was not possible 5 years ago. 
The objective of this report was to determine the availability and 
completeness of data that currently are collected on those 
experiencing homelessness, not on the extent to which these data have 
improved over time. In addition, HUD's data are not longitudinal in 
that they do not follow specific individuals over time; rather HUD 
collects aggregated data that track numbers of homeless over time. 
Nonetheless, in the report we discuss actions that HUD has taken to 
improve its homelessness data over time and note the inherent 
difficulties of collecting these data. The report also notes that 
HUD's point-in-time count represents the only effort by a federal 
agency to count all of those who are experiencing homelessness, rather 
than just those utilizing federally-funded programs. 

HUD made a number of other comments related to their data and the 
definition of homelessness. HUD commented that the report did not 
recognize that data collection is driven by statutory definitions or 
that HUD's point-in-time and HMIS systems are in some sense 
complementary. We have addressed this comment in the final report by 
making it clearer that data collected necessarily reflect the 
definitions included in the statutes that mandate data collection. We 
also added a footnote to show that while point-in-time counts focus on 
those who are homeless for long periods of time, HMIS may capture 
those who are homeless for shorter periods of time or move in and out 
of homelessness. HUD also commented that the report did not adequately 
describe the statutory history of homelessness definitions. We do not 
agree; the report describes the statutory history to the extent needed 
to address our objectives. Additionally, HUD commented that the report 
did not provide proper context about HMIS development and 
implementation at the local level, adding that a community's success 
in using HMIS to meet local needs depends on a variety of factors, 
such as staff experience and the quality of software selected. We 
revised the report to acknowledge that a community's success in using 
HMIS depends on these other factors. Further, the report acknowledges 
that in setting HMIS data standards, HUD allowed communities to adapt 
locally developed data systems or to choose from many other HMIS 
systems that meet HUD's standards. 

Finally, HUD wrote that we attribute the lack of collaboration among 
federal agencies solely to differences in definitions. Similarly, the 
Executive Director of the Interagency Council wrote that many greater 
obstacles to effective collaboration exist than the definitional 
issue--such as "siloed" departmental and agency structures, 
uncoordinated incentives and measures of effectiveness, difficulties 
communicating across very large bureaucracies, and different program 
rules for releasing and administering funds. The report does not 
attribute the lack of collaboration solely to the differences in 
definitions. Instead we note that agencies have not collaborated and 
that having a common vocabulary could improve collaboration. The 
report focuses on definitional differences, in part, because it was a 
key objective of our work and an issue frequently raised in 
discussions of barriers to effectively providing services to those 
experiencing homelessness. 

Education, HHS, and HUD also provided technical comments which we 
addressed as appropriate. 

We are sending copies of this report to the Secretaries of Education, 
Health and Human Services, Housing and Urban Development, Labor, and 
Justice; the Executive Director of the U.S. Interagency Council on 
Homelessness; and relevant congressional committees. This report will 
also be available at no charge on GAO's Web site at [hyperlink, 
http://www.gao.gov]. 

If you or your staff have any questions about this report, please 
contact me at (202) 512-8678 or cackleya@gao.gov. Contact points for 
our Offices of Congressional Relations and Public Affairs may be found 
on the last page of this report. GAO staff who made major 
contributions to this report are listed in appendix VII. 

Signed by: 

Alicia Puente Cackley: 
Director, Financial Markets and Community Investment: 

List of Requesters: 

The Honorable George Miller:
Chairman:
The Honorable John P. Kline:
Ranking Member:
Committee on Education and Labor:
House of Representatives: 

The Honorable Barney Frank:
Chairman:
The Honorable Spencer Bachus:
Ranking Member:
Committee on Financial Services:
House of Representatives: 

The Honorable Maxine Waters:
Chairwoman:
The Honorable Shelley Moore Capito:
Ranking Member:
Subcommittee on Housing and Community Opportunity:
Committee on Financial Services:
House of Representatives: 

The Honorable Judy Biggert:
Ranking Member:
Subcommittee on Oversight and Investigations:
Committee on Financial Services:
House of Representatives: 

The Honorable Geoff Davis:
House of Representatives: 

The Honorable Howard P. "Buck" McKeon:
House of Representatives: 

[End of section] 

Appendix I: Objectives, Scope and Methodology: 

The objectives of our report were to (1) assess the availability, 
completeness, and usefulness of data on homelessness collected by 
federal programs; (2) assess the extent to which research identifies 
factors associated with homelessness; and (3) analyze how differences 
in the definitions of homelessness and other factors, such as the 
level of agency collaboration, may impact the effectiveness of 
programs serving those experiencing homelessness. 

To address all of our objectives, we reviewed relevant laws such as 
the McKinney-Vento Homeless Assistance Act, as amended, and the HEARTH 
Act, as well as a range of prior GAO reports that addressed 
homelessness or related issues such as reviews of the Social Security 
Administration's Supplemental Security Income (SSI) and Supplemental 
Security Disability Income (SSDI) programs. We also reviewed 
regulations and government reports across a number of programs 
specifically targeted to address issues related to homelessness as 
well as mainstream programs, such as Temporary Assistance for Needy 
Families (TANF), Head Start, and Public Housing, that often provide 
services to people experiencing homelessness. Finally, we reviewed 
research on homelessness retrieved during a wide-ranging search of the 
literature. 

During our review, we conducted interviews with at least 60 entities, 
including officials of six federal government agencies, 
representatives of at least 15 state and local government entities, 
staff and officials at 27 service providers, 11 researchers, and 
officials at 10 groups that advocated for positions related to 
homelessness. These sum to more than the 60 interviews because some 
entities fall into more than one category. Specifically, we 
interviewed officials at the Departments of Education (Education), 
Health and Human Services (HHS), Housing and Urban Development (HUD), 
Justice (DOJ), and Labor (DOL), and the U.S. Interagency Council on 
Homelessness (Interagency Council). We also conducted in-depth 
interviews with advocates and researchers, as well as service 
providers, state and local government officials, and HUD field staff 
that had extensive experience with homeless programs. Many of our 
interviews were conducted as part of four site visits to large and 
medium-sized urban areas that were geographically distributed across 
the United States. We visited these locations to determine the extent 
to which views on homelessness were specific to particular locations 
or regions because of local laws, population concentration, or 
weather. We chose locations to represent each of the major regions of 
the United States--the Midwest, Northeast, South, and West--and to 
reflect differences in population concentration and weather. We chose 
specific urban areas in part because they had reported recent large 
changes in homelessness among families--two had seen a marked 
increase, while a third had noted a decrease. In the fourth location, 
homelessness had been relatively stable. Using these factors, we chose 
cities in California, Illinois, Massachusetts, and South Carolina. 
Generally, we did not consider issues specific to rural areas because 
Congress had mandated a separate study of them.[Footnote 49] We chose 
the specific organizations we interviewed to include a range of 
activities and views, but did not seek to interview a given number of 
agencies or individuals in each area or to develop a sample from which 
we could generalize our findings. 

We also undertook a number of activities specific to each objective: 

To address the first objective on the availability, completeness, and 
usefulness of data on homelessness collected by federal programs, we 
reviewed statutes, regulations, guidance, technical standards, and 
reports on federal data from targeted homelessness programs. We 
focused our review of federal data on the Housing and Urban 
Development Department's (HUD) Homeless Management Information System 
(HMIS) and point-in-time counts, Health and Human Services' (HHS) 
Runaway and Homeless Youth Management Information System (RHYMIS), and 
data submitted to the Department of Education through Consolidated 
State Performance Reports. We interviewed selected service providers 
to learn about the data systems they use to collect and store 
information on the homeless populations they serve, the procedures 
they use to ensure data reliability, and the usefulness of existing 
data systems for program management and administrative purposes. In 
addition, we interviewed selected federal, state, and local officials 
to identify the data used in their oversight of programs for families 
and individuals that are experiencing homelessness, the procedures 
they use to verify data reliability, and the extent to which existing 
data provide sufficient information for program management. Further, 
we spoke with researchers, individuals with special expertise with 
federal data systems, and government contractors, to determine the 
reliability and usefulness of existing data sources on the homeless, 
as well as to identify potential areas for improvement in data on the 
homeless. We also analyzed estimates of the extent of homelessness 
that were derived from federal data systems. In determining the 
reliability of the data for this report, we identified several 
limitations with the data-namely, that persons experiencing 
homelessness are hard to identify and count; that other than the point-
in-time count, the three federal data sources for targeted 
homelessness programs primarily capture data on program participants; 
and that duplication can exist because the population is mobile and 
dynamic-which are noted in the report. Nevertheless, because these are 
the only available data and the relevant departments use them to 
understand the extent and nature of homelessness, we present the data 
with their limitations. 

We also reviewed two HHS reports on homelessness and housing status 
data collected from federal mainstream programs, to determine the 
availability of such data. We reviewed research that estimated the 
size of the population that is doubled up with family and friends. We 
used data from the 2008 American Community Survey to develop our own 
estimate of the number of people who were experiencing severe to 
moderate economic hardship and living with an extended family or 
nonfamily member in 2008. The survey is conducted annually by the U.S. 
Census Bureau, and it asks respondents to provide information for 
housing information and employment income for households. We made 
several assumptions about what comprises severe or moderate economic 
hardship. Severe economic hardship was assumed to mean that households 
had housing costs of at least 50 percent of household income and that 
household income was below 50 percent of the federal poverty line and 
moderate economic hardship was assumed to mean that the households had 
housing costs that were at least 30 percent of household income and 
household income was below the federal poverty line. We also made 
assumptions about what comprises extended family; we assumed that 
extended family households were those where some people in the 
household were not part of the head of household's immediate family, 
and we included spouse, live-in partners, children, grandparents, and 
grandchildren in our definition of immediate family members. We cannot 
determine from the available data whether the individuals that are 
living with extended family or nonfamily members and experiencing 
severe or moderate economic hardship would meet the McKinney-Vento 
Children and Youth definition of homelessness, which requires that 
individuals be doubled up because of economic hardship. Because we 
followed a probability procedure based on random selections, our 
sample is only one of a large number of samples that we might have 
drawn. Since each sample could have provided different estimates, we 
express our confidence in the precision of our particular sample's 
results as a 95-percent confidence interval. This is the interval that 
would contain the actual population values for 95 percent of the 
samples we could have drawn. As a result, we cannot determine whether 
the people in our estimate would be eligible for the benefits if the 
McKinney-Vento Individual definition of homelessness were expanded to 
include those doubled up because of economic hardship. 

To address the second objective, we conducted a literature review to 
identify research studies that considered factors associated with the 
likelihood that families, youths, and individuals would experience 
homelessness. We also used various Internet search databases 
(including EconLit, ERIC, Medline, and Proquest) to identify studies 
published or issued after 1998. We chose 1998 as a starting point 
because welfare reform--which impacted some homeless families--had 
been implemented by that date and may have affected research findings. 
We sought to identify additional studies with persons we interviewed 
(that is, government officials, researchers, and advocacy groups) and 
from studies' bibliographies. In this initially broad search, we 
identified more than 600 studies, although we cannot be certain that 
we captured all relevant research that met our screening criteria. We 
screened the papers we identified using a multilevel process to gauge 
their relevance and evaluate their methodology. We excluded papers 
that did not specifically focus on our objective, were published or 
issued before 1998, lacked quantitative analysis, had a target 
population sample size of less than 25, did not conduct some form of 
statistical testing, did not use a comparison or control group or some 
other means to compare the target population (or group of persons to 
whom the research hopes to generalize findings) such as regression 
analyses, focused on homeless populations outside of the United 
States, or were dissertations. We retained 45 studies after screening 
and reviewed their methodologies, findings, and limitations. Nine GAO 
staff (four analysts and five methodologists) were involved in the 
systematic review of each of the 45 studies selected, which were 
determined to be sufficiently relevant and methodologically rigorous. 
More specifically, two staff members--one analyst and one 
methodologist--reviewed each study and reached agreement on the 
information entered in the database. 

As noted in this report, many of these studies are subject to certain 
methodological limitations, which may limit the extent to which the 
results can be generalized to larger populations. In some cases, 
studies did not discuss correlation among the factors and are thus 
limited in their ability to explain which factors might lead to 
homelessness. In addition, at least four studies used data that were 
more than 10 years old from the date of publication. Findings based on 
such data may be limited in explaining the characteristics and 
dynamics of current homeless populations. Further, collecting 
comparable information from individuals who have not been homeless (a 
comparison group) is important in determining which variables 
distinguish those experiencing homelessness from those that do not, 
and is essential in determining whether certain at-risk individuals 
and families experience homelessness and others do not. Although we 
generally excluded studies that did not use a comparison or control 
group to test their hypotheses, several studies in our literature 
review used a comparison group that was another homeless population 
rather than a nonhomeless control group. In addition to the literature 
review, we gathered opinions from researchers, advocates, service 
providers, and government officials on the factors associated with the 
likelihood of experiencing homelessness. 

To address the third objective, we took several steps to develop a 
list of potential barriers to providing services for those 
experiencing homelessness. First, we reviewed our prior work on 
barriers facing those experiencing homelessness. Second, we held 
initial interviews with researchers, service providers, and government 
officials in our Massachusetts location where potential barriers were 
raised. Third, in conjunction with a methodologist, we developed a 
list of potential barriers. The list, which included affordable 
housing, differences in definitions of homelessness used by various 
federal agencies, eligibility criteria other than income for accessing 
mainstream programs, the complexities of applying for grants, and lack 
of collaboration among federal agencies as well as a number of other 
potential barriers, was included in a structured data collection 
instrument to be used in the remaining interviews. We asked those we 
interviewed to select the three most important barriers from that list 
but did not ask them to rank order their selections. Interviewees were 
also able to choose barriers not on the list. To ensure that 
interviewees were interpreting the items on the list in the same way 
that we were interpreting them, we had interviewees describe the 
reasons for their choice. We determined the relative importance of the 
barriers chosen by summing the number of times an item was selected as 
one of the three most important barriers. 

When those we interviewed did not choose differences in definitions of 
homelessness as one of the three main barriers, we asked them for 
their views on definitional issues and asked all those we interviewed 
about the advantages of having multiple definitions of homelessness. 
Similarly, for collaboration among federal agencies, we asked those we 
interviewed about the agencies they worked with and, if they worked 
with multiple agencies, about their experiences. We also asked for 
examples of successful interaction among federal agencies. 
collaboration. As previously noted, lack of interagency collaboration 
was also on the list of barriers. In addition, we interviewed the 
acting and newly appointed executive directors of the Interagency 
Council on Homelessness and reviewed certain documents related to 
their activities; interviewed agency officials at Education, HUD, HHS, 
DOL, and DOJ; and reviewed agency planning and performance documents 
to identify coordination with other agencies. 

We conducted this performance audit from May 2009 to June 2010 in 
accordance with generally accepted government auditing standards. 
Those standards require that we plan and perform the audit to obtain 
sufficient, appropriate evidence to provide a reasonable basis for our 
findings and conclusions based on our audit objectives. We believe 
that the evidence obtained provides a reasonable basis for our 
findings and conclusions based on our audit objectives. 

[End of section] 

Appendix II: Bibliography of Studies GAO Reviewed That Analyze Factors 
Associated with Homelessness: 

We conducted a review of 45 research studies that analyzed factors 
associated with homelessness. Most of the studies we reviewed examined 
factors associated with the likelihood of entering an episode of 
homelessness or the rates of homelessness in a given area, while a few 
examined factors associated with the duration of homelessness. Twenty- 
nine studies examined adult individuals, 14 studies examined families, 
and 7 studies examined only youths.[Footnote 50] To assess factors 
associated with homelessness, studies used a range of analytical 
techniques--including measures of association or correlation between 
single factors and methods that accounted for some of the 
interrelationships among factors. The 45 studies are listed below: 

Allgood, Sam, and Ronald S. Warren, Jr. "The Duration of Homelessness: 
Evidence from a National Survey." Journal of Housing Economics 12 
(2003): 273-290. 

Anderson, Debra Gay, and M. K. Rayens. "Factors Influencing 
Homelessness in Women." Public Health Nursing 21, no. 1 (2004): 12-23. 

Bassuk, Ellen L., Jennifer N. Perloff, and Ree Dawson. "Multiply 
Homeless Families: The Insidious Impact of Violence." Housing Policy 
Debate 12 (2001): 299-320. 

Bendheim-Thoman Center for Research on Child Wellbeing and Columbia 
Population Research Center. "Predictors of Homelessness and Doubling-
up Among At-risk Families." Fragile Families Research Brief, no. 43 
(August 2008). 

Caton, Carol L. M., Boanerges Dominguez, Bella Schanzer, et al. "Risk 
Factors for Long-Term Homelessness: Findings from a Longitudinal Study 
of First-Time Homeless Single Adults." American Journal of Public 
Health 95 (2005): 1753-1759. 

Caton, Carol L. M., Deborah Hasin, Patrick E. Shrout, et al. "Risk 
Factors for Homelessness among Indigent Urban Adults with No History 
of Psychotic Illness: A Case-Control Study." American Journal of 
Public Health 90 (2000): 258-263. 

Collins, Cyleste C., Claudia J. Coulton, and Seok-Joo Kim. Family 
Homelessness in Cuyahoga County. White paper published for the Sisters 
of Charity Foundation, Center on Urban Poverty and Community 
Development, Cleveland, Ohio: Case Western Reserve University, 2009. 

Cousineau, Michael R. "Comparing Adults in Los Angeles County Who Have 
and Have Not Been Homeless." Journal of Community Psychology 296, no. 
6 (2001): 693-701. 

Culhane, Dennis P., and Stephen Metraux. "One-Year Rates of Public 
Shelter Utilization by Race/Ethnicity, Age, Sex and Poverty Status for 
New York City (1990 and 1995) and Philadelphia (1995)." Population 
Research and Policy Review (1999): 219-236. 

Culhane, Dennis P., Stephen Metraux, Stephen R. Poulin, and Lorlene M. 
Hoyt. "The Impact of Welfare Reform on Public Shelter Utilization in 
Philadelphia: A Time-Series Analysis." Cityscape: A Journal of Policy 
Development and Research, U.S. Department of Housing and Urban 
Development, Office of Policy Development and Research 6, no. 2 
(2003): 173-185. 

Early, Dirk W. "An Empirical Investigation of the Determinants of 
Street Homelessness." Journal of Housing Economics 14 (2005): 27-47. 

Early, Dirk W. "The Determinants of Homelessness and the Targeting of 
Housing Assistance." Journal of Urban Economics 55 (2004): 195-214. 

Early, Dirk W. "The Role of Subsidized Housing in Reducing 
Homelessness: An Empirical Investigation Using Micro-Data." Journal of 
Policy Analysis and Management 17, no. 4 (1998): 687-696. 

Eyrich-Garg, Karin M., John S. Cacciola, Deni Carise, et al. 
"Individual Characteristics of the Literally Homeless, Marginally 
Housed, and Impoverished in a U.S. Substance Abuse Treatment-Seeking 
Sample." Social Psychiatry and Psychiatric Epidemiology 43 (2008): 831-
842. 

Eyrich-Garg, Karin M., Catina Callahan O'Leary, and Linda B. Cottler. 
"Subjective Versus Objective Definitions of Homelessness: Are There 
Differences in Risk Factors among Heavy-Drinking Women?" Gender Issues 
25 (2008): 173-192. 

Fertig, Angela R., and David A. Reingold. "Homelessness among at-Risk 
Families with Children in Twenty American Cities." Social Service 
Review 82, no. 3 (2008): 485-510. 

Fitzgerald, Scott T., Mack C. Shelley II, and Paula W. Dail. "Research 
and Homelessness: Sources and Implications of Uncertainty." American 
Behavioral Scientist 45, no. 1 (2001): 121-148. 

Folsom, David P., William Hawthorne, Laurie Lindamer, et al. 
"Prevalence and Risk Factors for Homelessness and Utilization of 
Mental Health Services Among 10,340 Patients with Serious Mental 
Illness in a Large Public Mental Health System." American Journal of 
Psychiatry 162, no. 2 (2005): 370-376. 

Greenberg, Greg A., and Robert A. Rosenheck. "Homelessness in the 
State and Federal Prison Population." Criminal Behaviour and Mental 
Health 18, no. 2 (2008): 88-103. 

Gubits, Daniel, Jill Khadduri, and Jennifer Turnham. Housing Patterns 
of Low Income Families with Children: Further Analysis of Data from 
the Study of the Effects of Housing Vouchers on Welfare Families. 
Joint Center for Housing Studies of Harvard University, 2009. 

Ji, Eun-Gu. "A Study of the Structural Risk Factors of Homelessness in 
52 Metropolitan Areas in the United States." International Social Work 
49, no. 1 (2006): 107-117. 

Johnson, Timothy P., and Michael Fendrich. "Homelessness and Drug 
Use - Evidence from a Community Sample." American Journal of 
Preventive Medicine 32 (2007): S211-S218. 

Kingree, J. B., Torrance Stephens, Ronald Braithwaite, and James 
Griffin. "Predictors of Homelessness among Participants in a Substance 
Abuse Treatment Program." American Journal of Orthopsychiatry 69, no. 
2 (1999): 261-266. 

Kuhn, Randall, and Dennis P. Culhane. "Applying Cluster Analysis to 
Test a Typology of Homelessness by Pattern of Shelter Utilization: 
Results from the Analysis of Administrative Data." American Journal of 
Community Psychology 26 (1998): 207-232. 

Leal, Daniel, Marc Galanter, Helen Dermatis, and Laurence Westreich. 
"Correlates of Protracted Homelessness in a Sample of Dually Diagnosed 
Psychiatric Inpatients." Journal of Substance Abuse Treatment 16, no. 
2 (1999): 143-147. 

Lee, Barrett A., Townsand Price-Spratlen, and James W. Kanan. 
"Determinants of Homelessness in Metropolitan Areas." Journal of Urban 
Affairs 25 (2003): 335-355. 

Lehmann, Erika R., Philip H. Kass, Christiana M. Drake, and Sara B. 
Nichols. "Risk Factors for First-Time Homelessness in Low-Income 
Women." American Journal of Orthopsychiatry 77, no. 1 (2007): 20-28. 

Metraux, Stephen, and Dennis P. Culhane. "Family Dynamics, Housing, 
and Recurring Homelessness among Women in New York City Homeless 
Shelters." Journal of Family Issues 20, no. 3 (1999): 371-396. 

Molino, Alma C. "Characteristics of Help-Seeking Street Youth and Non- 
Street Youth." 2007 National Symposium on Homelessness Research, 2007. 

O'Flaherty, Brendan, and Ting Wu. "Fewer Subsidized Exits and a 
Recession: How New York City's Family Homeless Shelter Population 
Became Immense." Journal of Housing Economics (2006): 99-125. 

Olsen, Edgar O., and Dirk W. Early. "Subsidized Housing, Emergency 
Shelters, and Homelessness: An Empirical Investigation Using Data from 
the 1990 Census." Advances in Economic Analysis & Policy 2, no. 1 
(2002). 

Orwin, Robert G., Chris K. Scott, and Carlos Arieira. "Transitions 
through Homelessness and Factors That Predict Them: Three-Year 
Treatment Outcomes." Journal of Substance Abuse Treatment 28 (2005): 
S23-S39. 

Park, Jung Min, Stephen Metraux, and Dennis P. Culhane. "Childhood Out-
of-Home Placement and Dynamics of Public Shelter Utilization among 
Young Homeless Adults." Children and Youth Services Review 27, no. 5 
(2005): 533-546. 

Quigley, John M., Steven Raphael, and Eugene Smolensky. "Homeless in 
America, Homeless in California." The Review of Economics and 
Statistics 83, no. 1 (2001): 37-51. 

Rog, Debra J. C., Scott Holupka, and Lisa C. Patton. Characteristics 
and Dynamics of Homeless Families with Children. Final report to the 
Office of the Assistant Secretary for Planning and Evaluation, Office 
of Human Services Policy, U.S. Department of Health and Human Services 
(Rockville, Md.: Fall 2007). 

Shelton, Katherine H., Pamela J. Taylor, Adrian Bonner, and Marianne 
van den Bree. "Risk Factors for Homelessness: Evidence from a 
Population-Based Study." Psychiatric Services 60, no. 4 (2009): 465- 
472. 

Shinn, Marybeth, Beth C. Weitzman, Daniela Stojanovic, and James R. 
Knickman. "Predictors of Homelessness among Families in New York City: 
From Shelter Request to Housing Stability." American Journal of Public 
Health 88, no. 11 (1998): 1651-1657. 

Slesnick, Natasha, Suzanne Bartle-Haring, Pushpanjali Dashora, et al. 
"Predictors of Homelessness among Street Living Youth." Journal of 
Youth Adolescence 37 (2008): 465-474. 

Stein, Judith A., Michelle Burden Leslie, and Adeline Nyamathi. 
"Relative Contributions of Parent Substance Use and Childhood 
Maltreatment to Chronic Homelessness, Depression, and Substance Abuse 
Problems among Homeless Women: Mediating Roles of Self-Esteem and 
Abuse in Adulthood." Child Abuse & Neglect 26, no. 10 (2002): 1011-
1027. 

Sullivan, G., A. Burnam, and P. Koegel. "Pathways to Homelessness 
among the Mentally Ill." Social Psychiatry and Psychiatric 
Epidemiology 35 (2000): 444-450. 

Tyler, Kimberly A., and Bianca E. Bersani. "A Longitudinal Study of 
Early Adolescent Precursors to Running Away." Journal of Early 
Adolescence 28, no. 2 (2008): 230-251. 

The Urban Institute, Martha R. Burt, Laudan Y. Aron, et al. 
Homelessness: Programs and the People They Serve: Findings of the 
National Survey of Homeless Assistance Providers and Clients. 1999. 

Vera Institute of Justice, Nancy Smith, Zaire Dinzey Flores, et al. 
Understanding Family Homelessness in New York City: An In-Depth Study 
of Families' Experiences Before and After Shelter 2005. 

Whaley, Arthur L. "Demographic and Clinical Correlates of Homelessness 
among African Americans with Severe Mental Illness." Community Mental 
Health Journal 38, no. 4 (2002): 327-338. 

Yoder, Kevin A., Les B. Whitbeck, and Dan R. Hoyt. "Event History 
Analysis of Antecedents to Running Away from Home and Being on the 
Street." American Behavioral Scientist 45, no. 1 (2001): 51-65. 

[End of section] 

Appendix III: Comments from the Department of Education: 

United States Department Of Education: 
Office Of Elementary And Secondary Education: 
The Assistant Secretary: 
400 Maryland Ave. S.W. 
WASHINGTON, DC 20202: 
[hyperlink, http://www.ed.gov] 

The Department of Education's mission is to promote student 
achievement and preparation for global competitiveness by fostering 
educational excellence and ensuring equal access. 

June 18, 2010: 

Ms. Alicia Puente Cackley: 
Director, Financial Markets and Community Investment: 
U.S. Government Accountability Office: 
441 G Street, NW: 
Washington, DC 20548: 

Dear Ms. Cackley: 

I am writing in response to your request for comments on the U.S. 
Government Accountability Office's (GAO's) draft report (GAO 10-702) 
dated June 2010, entitled "Homelessness: A Common Vocabulary Could 
Help Agencies Collaborate and Collect More Consistent Data." I 
appreciate the opportunity to comment on the draft report on behalf of 
the Department of Education. 

The following are the Department's responses to the report's 
recommendations for executive action. 

Recommendation 1: Develop joint federal guidance that establishes a 
common vocabulary for discussing homelessness and related terms. Such 
guidance may allow these and other agencies on the Interagency Council 
on Homelessness to collaborate more effectively to provide coordinated 
services to those experiencing homelessness. 

The Department agrees that some coordination of services for homeless 
children and youth across federal agencies with programs serving this 
population is helpful, and recognizes that some consideration of 
common vocabulary is appropriate to help provide useful guidance. We 
look forward to continuing to work with the U.S. Interagency Council 
on Homelessness (USICH) on these matters. The Department and other 
federal agencies have begun discussions about our definitions of 
homelessness as part of preparing the forthcoming Federal Strategic 
Plan to Prevent and End Homelessness. To work toward common 
definitions and terminology would, in some cases, require 
congressional action to make statutory conforming changes. In 
considering these matters, the Department will help ensure that any 
common definition would not undermine the intent of the Education for 
Homeless Children and Youth program (that homeless children and youth 
enroll, attend, and succeed in school) and that the effectiveness of 
the program in serving homeless children is not reduced. 

Recommendation 2: Determine whether the benefits of using this common 
vocabulary to develop and implement guidance for collecting consistent 
federal data on housing status for targeted homelessness programs as 
well as mainstream programs that address the needs of low income 
populations would exceed the costs. 

Depending on the consideration of the actions related to 
Recommendation 1, the Department proposes that a discussion of the 
costs and benefits of using more of a common vocabulary, as it relates 
to data collection, be an agenda item for USICH. 

We appreciate the opportunity to review the draft report and comment 
on the recommendations. Overall, the investigating team prepared a 
comprehensive and accurate report of the perspectives of people 
affiliated with the Department's targeted program for the education of 
homeless children and youth in various regions of the country and 
levels of government, as well as other program stakeholders. I am also 
enclosing our technical comments. 

Sincerely, 

Signed by: 

Thelma Melendez de Santa Ana, Ph.D. 
Assistant Secretary: 
Office of Elementary and Secondary Education: 

[End of section] 

Appendix IV: Comments from the Department of Health and Human Services: 

Department Of Health & Human Services: 
Office of the Assistant Secretary for Legislation: 
Washington, D.C. 20201: 

June 23, 2010: 

Alicia Puente Cackley, Director: 
Financial Markets and Community Investment: 
U.S. Government Accountability Office: 
441 G Street N.W. 
Washington, DC 20548: 

Dear Ms. Cackley: 

Attached are the Department's comments on the U.S. Government 
Accountability Office's (GAO) report entitled: "HOMELESSNESS: A Common 
Vocabulary Could Help Agencies Collaborate and Collect More Consistent 
Data" (GAO-10-702). 

The Department appreciates the opportunity to review this report 
before its publication. 

Sincerely, 

Signed by: 

Andrea Palm: 
Acting Assistant Secretary for Legislation: 

[End of letter] 

General Comments Of The Department Of Health And Human Services (HHS) 
On The Government Accountability Office's Draft Report Entitled, 
"Homelessness: A Common Vocabulary Could Help Agencies Collaborate And 
Collect More Consistent Data" (GAO-10-702): 

The Department appreciates the opportunity to comment on this GAO 
draft report. 

GAO Recommendations: 

To help improve their understanding of homelessness and to help 
mitigate the barriers posed by having differences in definitions of 
homelessness and related terminology, we recommend that the 
Secretaries of Education, HHS, and HUD---working through the U.S. 
Interagency Council on Homelessness: 

1. Develop joint federal guidance that establishes a common vocabulary 
for discussing homeless and related terms. Such guidance may allow 
these and other agencies on the Interagency Council on Homelessness to 
collaborate more effectively to provide coordinated services to those 
experiencing homelessness; and; 

2. Determine whether the benefits of using this common vocabulary to 
develop and implement guidance for collecting consistent federal data 
on housing status for targeted homelessness programs as well as 
mainstream programs that address the needs of low income populations 
would exceed the costs. 

HHS Comments: 

The report addresses a set of perennial and interrelated issues in 
measuring and understanding the problem(s) of homelessness. These 
include the multiple definitions of "homelessness," consistency/non-
duplication of data on homeless individuals and families, and the 
patchwork mixture of data systems collecting data on these 
populations. However, the report does not acknowledge that there can 
be a sound rationale for using different definitions of homelessness. 

Pure definitional uniformity per se across the wide range of programs 
for access to housing, education, mental health services, income 
assistance, and emergency shelter (for adults and for minors) is not 
necessarily a desirable goal. Different populations, different 
statutory mandates, and different frames of reference require 
specialized terminology so that services are effectively targeted and 
client groups with variations are well-served. The specific 
definitions that programs use may help some programs better target 
their resources on the unique needs and circumstances of the 
populations they are designed to serve. The adult homelessness system 
is not the best model for child homelessness and is definitely not an 
appropriate venue for shelter or services, certainly not for children 
under 18 who are by themselves. 

For example, on Page 2, GAO points out the IRIS' Runaway and Homeless 
Youth programs define homelessness differently than other programs 
because it includes lacking a safe living environment as a criterion. 
It can be argued that a homeless definition that includes safety as a 
criterion is appropriate for runaway and homeless youth because this 
population is young, vulnerable and particularly at risk. 

In many ways, the Administration for Children and Familes' (ACF) 
Family and Youth Services Bureau (FYSB) Runaway and Homeless Youth 
Programs (RHY), particularly the Basic Centers, which receive the 
largest share of RHY funding, are primarily child protection, family 
reunification and/or appropriate placement systems. Basic Centers 
serve youth until they become 18. The Transitional Living Program 
(TLP) for older homeless youth resembles adult transitional programs 
in many ways, but its client base is an order of magnitude smaller 
than the Basic Center caseload, and well over one half of the TLP 
caseload is aged only 16 through 18 years. 

Unaccompanied youth/child homelessness is very different from adult or 
family homelessness. 

* Young people are still developing physically and emotionally. 

* Children are even more vulnerable to exposure than adult homeless 
persons and may have become homeless traumatically because of abuse, 
neglect or abandonment, as well as because of the economic, employment 
and housing issues which drive adult homelessness. Children and young 
people are unquestionably less sturdy than adults or intact family 
groups. 

* Mental health and substance abuse play roles in both youth and adult 
homelessness, but substance abuse is known to have a much harsher 
impact on adolescents and their development than on adults (except, of 
course, where such adults began their substance abuse while children). 

* Once on the streets, youth are targets for exploitation and often 
must engage in prostitution or petty crime in order to survive. This 
is also true of adults, but society recognizes the vast differences in 
strength and cunning between youth and adults. 

HHS notes that GAO recognized the ability of its Runaway and Homeless 
Youth Management Information System (RHYMIS) to collect unduplicated 
counts of youth served in runaway shelters and homeless youth 
transitional residencies no matter where in the United States they may 
be housed and served. Non-duplication is achievable within a given 
continuum but not always when individuals move from one system to 
another in a different geographical area. 

GAO also notes that RHYMIS facilitates the rollup of a large set of 
variables into non-duplicative national or regional aggregates, and 
its network of providers do frequently (every six months). Moreover, 
the data are validated and available only a few months after each 
transfer. 

GAO recognized that FYSB has provided guidance and technical 
assistance for software developers endeavoring to link HMIS and 
RHYMIS. However, the report did not note that assuring such customized 
systems can transfer data relatively seamlessly into RHYMIS is 
expensive and time-consuming. The report states that the HHS has not 
provided funding for these endeavors (page 18), but in fact, FYSB has 
devoted a considerable share of its resources to this end, resources 
that are limited by strict limitations on what proportion of the 
appropriation may be devoted to efforts other than funding local RHY 
programs. To devote more resources to RHYMIS would inevitably mean 
less resources for other components of the RHY "network of support," 
for example, the National Runaway Switchboard, a vital lifeline for 
runaways or potential runaways looking for help in critical situations. 

Finally, the potential emergence of a patchwork of locally improvised 
HMIS/RHYMIS systems could greatly diminish the amount of reliable, 
timely, detailed, and nationally consistent data that RHYMIS has been 
able to provide at a response rate of between 96 and 99 percent year 
after year. Ironically, systems have been developed for private 
systems to link with HMIS, but these became useless when continua 
suddenly switched their HMIS software, negating the efforts undertaken 
(page 18). 

In the report's discussion of a literature review of 45 research 
studies analyzing factors associated with homelessness, GAO expresses 
concern that these studies are deficient because they used varying 
definitions for concepts like homelessness, youth and family 
relationships. This section fails to acknowledge that social science 
research has many examples where terminology varies and yet this does 
not impede researchers from looking across a range of such studies to 
do comparative analyses and synthesize findings from a body of 
research. Such research has informed us about the characteristics and 
needs of homeless populations, and has identified promising or 
effective interventions for some segments of homeless populations. 

Concerning data collection for the National Youth in Transition 
Database, GAO states: "Because of staff turnover at HHS and 
difficulties with state data systems, it has taken over 10 years to 
implement the survey." 

As background, the Chafee Foster Care Independence Act (P.L. 106-169), 
enacted in December 1999, mandated the collection of data from States 
on the number and characteristics of children receiving services, as 
well as outcome measures that could be used to assess States' 
performance in operating independent living programs. 

Following the specifications mandated in the statute, ACF's Children's 
Bureau reviewed existing child welfare literature to identify data 
elements to be collected and reported in the database, and engaged in 
extensive consultation with: child welfare administrators and youth 
services coordinators at State, Tribal, and local levels; public and 
private agency youth services providers; child welfare advocates; 
group home staff and administrators; and current and former foster 
youth and foster parents. The Children's Bureau also established a 
pilot test in seven states to assess burden for workers collecting 
data. 

In September 2001, HHS submitted a report to Congress outlining a plan 
and timetable for developing this data collection system, and 
indicated that to maximize input from States and stakeholders, we 
would be publishing proposed and final outcomes and measures in a 
proposed rule rather than produce Program Instructions, as initially 
planned. 

Following this report, the Children's Bureau analyzed the results of 
the pilot test and consulted with the Standing Work Group of national 
experts to develop proposed data elements, instruments, and an 
implementation plan for the National Youth in Transition Database 
(NYTD). The Children's Bureau also developed a proposal to include a 
penalty structure for those states who failed to comply with NYTD 
requirements. 

In July 2006, the NYTD Notice of Proposed Rulemaking (NPRM) was 
published in the Federal Register, which detailed scope and reporting 
populations, data elements, data collection schedule, implementation 
plan, and penalty structure. The proposed rules initiated a 
substantial series of interim steps necessary to implement the 
database including a 60-day comment period, collection and analysis of 
public comments, full agency and department review, and Office of 
Management and Budget (OMB) clearance. 

The NYTD Final Rule was published in the Federal Register on February 
26, 2008, citing a NYTD implementation date of October 1, 2010, to 
provide states with sufficient lead time to develop/modify their own 
data systems, develop data collection instruments and methods, and 
acquire the necessary training/technical assistance to successfully 
collect and submit NYTD data. 

In the "Conclusions" section and elsewhere in the report, GAO appears 
to assume that programs identify people who are homeless only to have 
a total count of the homeless population. For example, on Page 46, a 
distinction is not made between identifying program participants who 
are homeless so that they can be referred for housing assistance, and 
identifying the extent and nature of overall homelessness in the 
population. Here, the report appears to imply that one of the roles of 
agencies delivering services to homeless populations is to collect 
data to determine the full extent and nature of homelessness, when few 
programs would have the authority or resources to carry out this 
function. 

[End of section] 

Appendix V: Comments from the Department of Housing and Urban 
Development: 

U.S. Department Of Housing And Urban Development: 
Assistant Secretary For Community Planning And Development: 
WASHINGTON, DC 20410-7000: 

June 22, 2010: 

Ms. Alicia Cackley, Director: 
Financial Markets and Community Investment: 
U.S. Government Accountability Office: 
441 G Street, NW, Room 2017: 
Washington, DC 20548: 

Dear Ms. Cackley: 

This is in response to the Government Accountability Office's draft 
report entitled Homelessness A Common Vocabulary Could Help Agencies 
Collaborate and Collect More Consistent Data (June 2010)—GAO-10-702. 
On June 21, 2010, HUD provided extensive comments under separate 
cover, but wishes to express its disappointment with the report. While 
HUD agrees with the overall GAO recommendation regarding the value of 
a common vocabulary. HUD found that the report does not adequately 
describe or address complex issues related to the definition of 
homelessness and data collection. 

Overall the report presents an incomplete and, therefore, unbalanced 
review of HUD's data collection and reporting efforts and does not 
attempt to recognize the great strides that have been made in this 
area or the value of the data currently being collected or reported. 
Below is an overview of HUD's concerns: 

* The report dedicates only three sentences to the statutory history 
of the definition of homelessness and does not adequately describe the 
changes authorized in the three revisions to the definition of 
homelessness within the McKinney-Vento Act in relation to each other. 

* The report does not emphasize that the data collected by each agency 
are derived from the definition of homelessness the program is 
statutorily required to use. 

* The report does not reflect the significant progress made by HUD in 
working with communities to collect comprehensive, accurate data on 
the extent and nature of homelessness. The report also does not 
acknowledge that HUD's Annual Homeless Assessment Report (AHAR) is the 
only national report on homelessness to use longitudinal data. 

* The report addresses the limitations of the Point-In-Time count and 
Homeless Management Information System (HMIS), but does not present an 
assessment on how the two data collection methods work in conjunction 
to address shortfalls. 

* The report does not provide proper context about HMIS development 
and implementation at the local level. A community's success in using 
HMIS to meet local needs depends on a variety of factors such as staff 
experience, the quality of the HMIS software selected, and the 
existence of other data systems. 

Ultimately, the report attributes the inability of federal agencies to 
collaborate effectively exclusively to definitional differences. In 
many instances, federal agencies are unable to collaborate due to 
authorizing statutes and implementing regulations. While a common 
vocabulary would improve the understanding of homelessness and allow 
for more informed local and federal discussions, a common vocabulary 
will not mitigate all of the barriers to serving homeless persons 
using various funding streams. 

HUD appreciates the opportunity to respond to this report and would 
welcome the opportunity to discuss these concerns. 

Sincerely, 

Signed by: 

Mercedes Marquez: 
Assistant Secretary: 

[End of section] 

Appendix VI: Comments from the Executive Director of the U.S. 
Interagency Council on Homelessness: 

United States Interagency Council on Homelessness: 

June 21, 2010: 

Dear Ms. Cackley: 

We at the United States Interagency Council on Homelessness (USICH) 
appreciate the opportunity to comment on the GAO report: Homelessness: 
A Common Vocabulary Could Help Agencies Collaborate and Collect More 
Consistent Data." 

First, lack of programmatic success or positive outcomes for those 
experiencing homelessness are not solely attributable to the varied 
definitions. There are many greater obstacles to effective 
collaboration than the definitional issue, including siloed 
departmental and agency structures, uncoordinated incentives and 
measurement of effectiveness, difficulties communicating across very 
large bureaucracies, and different program rules for releasing and 
administering funds. The report seems to give disproportionate weight 
to the definition. 

As the federal entity responsible for interagency coordination, USICH 
is actively engaged in creating bridges between agencies and the 
programs they administer in order to best assist those who are 
experiencing homelessness. In "Opening Doors: The Federal Strategic 
Plan to Prevent and End Homelessness" which was transmitted to the 
Congress and the President on June 22", USICH addressed the issue of 
definitions as follows: 

"The Plan acknowledges and supports the full range of federal 
definitions of homelessness as prescribed in statute, as each plays an 
appropriate and essential role in supporting and stabilizing those 
whom they are intended to help...If we are to truly end homelessness, 
we must use all resources that exist—both those that are intended for 
targeted homeless populations and those that are available for a 
broader segment of the population—to create lasting bridges across 
current gaps in housing and services." 

Second, there have been tremendous strides forward in the collection 
of data on homelessness, especially through the Annual Homeless 
Assessment Report completed by the Department of Housing and Urban 
Development. The report focuses on the limitations and short-comings, 
but does not note what is possible today that was not possible even 
five years ago. 

Third, the report is accurate on the limitations of existing research. 
We agree that most studies have been local and that there has not been 
much attention paid to how to identify which interventions will work 
best for which populations in order to most efficiently prevent and 
end homelessness. 

Finally, the report accurately documents the benefits of collecting 
data on housing status across mainstream programs and the barriers to 
doing that cost effectively. USICH supports further exploration of how 
to accurately and consistently report housing status in mainstream 
programs and supports the use of a "common vocabulary" to describe 
housing status that would improve the quality of data that is 
collected. The Homelessness Management Information System (HMIS) 
definitions of housing status reflect considerable work in this area 
and should be considered for adoption by other programs. 

Thank you for the opportunity to comment. 

Signed by: 

Barbara Poppe: 
Executive Director: 
United States Interagency Council on Homelessness: 

[End of section] 

Appendix VII: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

Alicia Puente Cackley, (202) 512-8678, or cackleya@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, Paul Schmidt, Assistant 
Director; Nancy S. Barry; Katie Boggs; Russell Burnett; William 
Chatlos; Kimberly Cutright; Marc Molino; Barbara Roesmann; Paul 
Thompson; Monique Williams; and Bryan Woliner made major contributions 
to this report. 

[End of section] 

Footnotes: 

[1] The HEARTH Act is set forth in Division B of Pub. L. No. 111-22 § 
1001, et seq. (May 20, 2009). The effective date of the HEARTH Act 
amendments to McKinney-Vento is the earlier of 18 months from the date 
of enactment (May 20, 2009) or 3 months after the publication of HUD's 
final regulations implementing the amendments, which are to be 
promulgated not later than 12 months after the date of enactment. Pub. 
L. No. 111-22 § 1503. Unless otherwise indicated, the McKinney-Vento 
Act definitions discussed in this report are those in effect before 
the effective date. Pursuant to the Hearth Act, HUD recently published 
a rule proposal to clarify key terms in the Act's definitions of 
"homeless," "homeless individual," "homeless person," and "homeless 
individual with a disability." 75 Fed. Reg. 20541 (Apr. 20, 2010). 

[2] GAO, Homelessness: Coordination and Evaluation of Programs Are 
Essential, [hyperlink, http://www.gao.gov/products/GAO/RCED-99-49] 
(Washington, D.C.: Feb. 26, 1999); Homelessness: Barriers to Using 
Mainstream Programs, [hyperlink, 
http://www.gao.gov/products/GAO/RCED-00-184] (July 6, 2000); and 
Homelessness: Improving Coordination and Client Access to Programs, 
[hyperlink, http://www.gao.gov/products/GAO-02-485T] (Washington, 
D.C.: Mar. 6, 2002). 

[3] The act was originally named the Stewart B. McKinney Act but was 
changed to the McKinney-Vento Act in 1989. 

[4] Prior to the inclusion of this definition in the McKinney-Vento 
education subtitle in 2002, similar language was contained in policy 
guidance issued by Education in 1995. 

[5] Pub. L. No. 100-77 § 103, as amended, 42 U.S.C. § 11302 (2008). 

[6] 42 U.S.C. § 11434a (2008). See "McKinney-Vento Homeless Education 
Assistance Improvements Act of 2001," Pub. L. 107-110 §§ 1031, 1032, 
115 Stat. 1989, 2005-06 (Jan. 8, 2002). 

[7] For Transitional Housing and Supportive Services Only projects, 
HUD recently extended the residency requirement for a stay in an 
institution to 90 days or more. HUD Notice of Funding Availability, 74 
Fed. Reg. 50816 (Oct. 1, 2009). 

[8] For additional changes to the general definition of homelessness 
in the HEARTH Act, see Pub. L. No. 111-22 § 1003(a). 

[9] Pub. L. No. 111-22 § 1003(a). 

[10] Runaway and Homeless Youth age eligibility requirements vary by 
program. Basic Center programs serve youths under 18. Transitional 
Living Programs serve youths from age 16 to age 22, unless an 
individual was admitted to the program before reaching 22 years of age 
and has not exceeded the maximum stay in the program by age 22. See 42 
U.S.C. § 5732a. The Runaway and Homeless Youth Act (Pub. L. No. 93-
415, title III (Sept 7, 1974)) was most recently reauthorized by the 
Reconnecting Homeless Youth Act of 2008 (P.L. 110-378). 

[11] The Council was established by title II of the McKinney-Vento 
Act, Pub. L. No. 100-77 § 20, as the "Interagency Council on the 
Homeless." In 2004, Congress renamed it "United States Interagency 
Council on Homelessness." Pub. L. No. 108-199 § 201 (Jan. 23, 2004). 

[12] See Pub. L. No. 111-22 § 1004. 

[13] VA collects data on veterans experiencing homelessness through an 
annual survey of VA staff, service providers, and veterans currently 
or formerly experiencing homelessness. The survey is used to catalogue 
the needs of these veterans and estimate the size of their population. 
HUD plans to issue a Veteran's Supplement to their next annual report 
on homelessness to Congress, which will focus exclusively on veterans. 

[14] See Pub. L. No. 106-377, Appendix A, 114 Stat 1441A-18 (Oct 27, 
2000). 

[15] For a discussion of the congressional directive, see Report to 
Congress: HUD's Strategy for Homeless Data Collection, Analysis and 
Reporting (August 2001), available at Congressional Directive/HUD 
Study, [hyperlink, 
http://www.hud.gov/offices/cpd/homeless/hmis/strategy/]. 

[16] The Census Bureau analyzed data from the 2000 Census to report on 
the number of people in emergency and transitional shelters, but due 
to data limitations, the Census Bureau advises against using the data 
as a count of the population experiencing homelessness. See U.S. 
Census Bureau, Emergency and Transitional Shelter Population: 2000 
(Washington, D.C.: U.S. Government Printing Office, 2001). 

[17] HUD also uses data submitted by Continuums through HMIS that 
capture some of the persons experiencing short term periods of 
homelessness or persons who periodically cycle in and out of 
homelessness. 

[18] See Violence Against Women and Department of Justice 
Reauthorization Act of 2005, Pub. L. No. 109-261 § 605 (Jan 5, 2005); 
see also, The Violence Against Women and Department of Justice 
Reauthorization Act of 2005: Applicability to HUD Programs; Notice, 72 
Fed. Reg. 12696 (March 16, 2007). 

[19] As previously discussed, HUD requires communities to conduct 
counts biennially. However, 67 percent of Continuums voluntarily 
conducted a count in 2008, a year in which HUD did not require a 
count. HUD uses data from communities that do submit counts every year 
to estimate the number of homeless in years where a count is not 
required. 

[20] For the purposes of estimating annual homelessness, HUD samples 
Community Development Block Grant entitlement communities. Communities 
not included in the statistical sample have also submitted data. 

[21] These estimates account for sheltered homeless people who used an 
emergency shelter or a transitional housing program at any time from 
October 1, 2007, through September 30, 2008. The estimates do not 
account for (1) persons who lived on the streets or in places not 
meant for human habitation, or who did not access a residential 
homeless program during the 1-year reporting period and (2) persons 
who used only a domestic violence shelter and did not access a 
residential homeless program that serves the general population. 

[22] In its comments on this report, HHS provided a detailed history 
of the development of the National Youth in Transition Database. See 
appendix IV for this information. 

[23] HHS, Homeless Data in Health and Human Services Mainstream 
Programs (Washington, D.C.: Winter 2009). 

[24] See appendixes I and II for more information on our criteria for 
selecting the studies, the methodology for reviewing them, and a list 
of the studies we reviewed. 

[25] Bendheim-Thoman Center for Research on Child Wellbeing and 
Columbia Population Research Center, "Predictors of Homelessness and 
Doubling-up among At-risk Families," Fragile Families Research Brief, 
no. 43 (August 2008) and Cyleste C. Collins, Claudia J. Coulton, and 
Seok-Joo Kim, Family Homelessness in Cuyahoga County, white paper 
published for the Sisters of Charity Foundation, Center on Urban 
Poverty and Community Development (Cleveland, Ohio: Case Western 
Reserve University, 2009). 

[26] HUD's 2010 Open Government Plan outlines an initiative to utilize 
federal data to build predictive models of homelessness. 

[27] The Fragile Families and Child Wellbeing Study follows a group of 
nearly 5,000 children born in large U.S. cities from 1998 to 2000 
(roughly three-quarters of whom were born to unmarried parents). The 
study consists of interviews with both mother and father at their 
child's birth and again when children are ages one, three and five, 
plus in-home assessments of children and their home environments at 
ages three and five. The data were collected primarily to address the 
conditions and capabilities of unmarried parents; relationships 
between unmarried parents; how children in these families fare; and 
how policies and environmental conditions affect families and 
children. DOL funds the National Longitudinal Survey of Youth, which 
began in 1979 to survey cohorts of 14 to 22 year olds annually. Survey 
questions focus on economic, social, and academic experiences of 
respondents and examine the complex issues surrounding youths entry 
into the work force and subsequent transitions in and out of it. 

[28] Natasha Slesnick, Suzanne Bartle-Haring, Pushpanjali Dashora, et 
al., "Predictors of Homelessness among Street Living Youth," Journal 
of Youth Adolescence 37 (2008): 465-474 and Arthur L. Whaley, 
"Demographic and Clinical Correlates of Homelessness among African 
Americans with Severe Mental Illness," Community Mental Health Journal 
38, no. 4 (2002), 327-338. 

[29] Bendheim-Thoman Center, 2008. 

[30] Debra J. C. Rog, Scott Holupka, and Lisa C. Patton, 
Characteristics and Dynamics of Homeless Families with Children. Final 
Report, final report to the Office of the Assistant Secretary for 
Planning and Evaluation, Office of Human Services Policy, U.S. 
Department of Health and Human Services (Washington, D.C.: Fall, 2007). 

[31] John M. Quigley, Steven Raphael, and Eugene Smolensky, "Homeless 
in America, Homeless in California." Review of Economics and 
Statistics 83, no. 1 (2001), 37-51. 

[32] Eun-Gu Ji, "A Study of the Structural Risk Factors of 
Homelessness in 52 Metropolitan Areas in the United States," 
International Social Work 49 (2006), 107-117. 

[33] As noted previously, most of these relied on experiences of 
childhood reported by adults experiencing homelessness. 

[34] Carol L Caton, M. D. Hasin, Patrick E. Shrout, et al., "Risk 
Factors for Homelessness among Indigent Urban Adults with No History 
of Psychotic Illness: A Case-Control Study," American Journal of 
Public Health 90 (2000), 258-263. 

[35] Daniel Gubits, Jill Khadduri, and Jennifer Turnham, Housing 
Patterns of Low Income Families with Children: Further Analysis from 
the Study of the Effects of Housing Vouchers on Welfare Families, 
Joint Center for Housing Studies of Harvard, 2009 and Erika R. 
Lehmann, Philip H. Kass, Christiana M. Drake, and Sara B. Nichols, 
"Risk Factors for First-Time Homelessness in Low-Income Women," 
American Journal of Orthopsychiatry 77, no. 1 (2007), 20-28. 

[36] Karin M. Eyrich-Garg, Catina Callahan O'Leary, and Linda B. 
Cottler, "Subjective Versus Objective Definitions of Homelessness: Are 
There Differences in Risk Factors among Heavy-Drinking Women?" Gender 
Issues 25, no. 3 (2008), 173-192. 

[37] Bendheim-Thoman Center, 2008. 

[38] Angela R. Fertig and David A. Reingold, "Homelessness among at- 
Risk Families with Children in Twenty American Cities," Social Service 
Review 82, no. 3 (2008), 485-510. 

[39] Marjorie Robertson, et al., "Rural Homelessness," Toward 
Understanding Homelessness: The 2007 National Symposium on 
Homelessness Research, Deborah Dennis, Gretchen Locke, and Jill 
Khadduri, ed. (Washington, D.C.: September 2007). 

[40] Before enactment of the HEARTH Act, there was no statutory 
definition of chronic homelessness. In a 2003 Federal Register release 
announcing a joint agency initiative to end chronic homelessness, HUD, 
HHS and VA defined the term to mean "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." Notice of Funding Availability 
for the Collaborative Initiative to Help End Chronic Homelessness, 68 
Fed. Reg. 4018, 4019 (Jan. 27, 2003). In an amendment to title IV of 
the McKinney-Vento Act (which authorizes the emergency shelter and 
transitional housing programs administered by HUD), the HEARTH Act 
added a definition of "chronically homeless" that includes individuals 
or families who qualify under the definition. Pub. L. No. 111-22 § 
1101, 123 Stat. 1669. 

[41] As previously noted, HUD recently extended the residency 
requirement for a stay in an institution for the Transitional Housing 
and Supportive Services Only programs to 90 days or more. The HEARTH 
Act also defines chronic homelessness to include those in an 
institution for less than 90 days that had previously met the 
definition of chronic homelessness. 

[42] Randall Kuhn and Dennis P. Culhane, "Applying Cluster Analysis to 
Test a Typology of Homelessness by Pattern of Shelter Utilization: 
Results from the Analysis of Administrative Data," American Journal of 
Community Psychology 26 (1998) 207-32. 

[43] Pub. L. No. 111-22 § 1004(a)(2). Congress also mandated that the 
Interagency Council consider barriers to serving those experiencing 
homelessness and how differences in definitions create barriers to 
collaboration among federal agencies during this convening of experts. 

[44] [hyperlink, http://www.gao.gov/products/GAO/RCED-00-184]. 

[45] GAO, Results-Oriented Government: Practices That Can Help Enhance 
and Sustain Collaboration among Federal Agencies, [hyperlink, 
http://www.gao.gov/products/GAO-06-15] (Washington, D.C.: Oct. 21, 
2005). 

[46] [hyperlink, http://www.gao.gov/products/GAO/RCED-99-49] and 
[hyperlink, http://www.gao.gov/products/GAO-02-485T]. 

[47] The strategic plan, U.S. Interagency Council on Homelessness, 
Opening Doors: Federal Strategic Plan to Prevent and End Homelessness 
(Washington, D.C.: June 2010), was not issued early enough to be fully 
considered in this report. 

[48] HUD's 2011 budget proposal does not include funding for new HUD- 
VASH vouchers. 

[49] The HEARTH Act mandated that we conduct a study to examine 
homelessness and homeless assistance in rural areas and rural 
communities. A report based on that study will be issued in July 2010. 

[50] Two studies examined a combination of adults and youths, five 
examined both adult individuals and families, and one examined a 
combination of adults, families, and youths. 

[End of section] 

GAO's Mission: 

The Government Accountability Office, the audit, evaluation and 
investigative arm of Congress, exists to support Congress in meeting 
its constitutional responsibilities and to help improve the performance 
and accountability of the federal government for the American people. 
GAO examines the use of public funds; evaluates federal programs and 
policies; and provides analyses, recommendations, and other assistance 
to help Congress make informed oversight, policy, and funding 
decisions. GAO's commitment to good government is reflected in its core 
values of accountability, integrity, and reliability. 

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through GAO's Web site [hyperlink, http://www.gao.gov]. Each 
weekday, GAO posts newly released reports, testimony, and 
correspondence on its Web site. To have GAO e-mail you a list of newly 
posted products every afternoon, go to [hyperlink, http://www.gao.gov] 
and select "E-mail Updates." 

Order by Phone: 

The price of each GAO publication reflects GAO’s actual cost of
production and distribution and depends on the number of pages in the
publication and whether the publication is printed in color or black and
white. Pricing and ordering information is posted on GAO’s Web site, 
[hyperlink, http://www.gao.gov/ordering.htm]. 

Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537. 

Orders may be paid for using American Express, Discover Card,
MasterCard, Visa, check, or money order. Call for additional 
information. 

To Report Fraud, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: [hyperlink, http://www.gao.gov/fraudnet/fraudnet.htm]: 
E-mail: fraudnet@gao.gov: 
Automated answering system: (800) 424-5454 or (202) 512-7470: 

Congressional Relations: 

Ralph Dawn, Managing Director, dawnr@gao.gov: 
(202) 512-4400: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7125: 
Washington, D.C. 20548: 

Public Affairs: 

Chuck Young, Managing Director, youngc1@gao.gov: 
(202) 512-4800: 
U.S. Government Accountability Office: 
441 G Street NW, Room 7149: 
Washington, D.C. 20548: