Homeless Veterans Programs: Improved Communications and Follow-up
Could Further Enhance the Grant and Per Diem Program (11-SEP-06, 
GAO-06-859).							 
                                                                 
About one-third of the nation's adult homeless population are	 
veterans, according to the Department of Veterans Affairs (VA).  
Many of these veterans have experienced substance abuse, mental  
illness, or both. The VA's Homeless Providers Grant and Per Diem 
(GPD) program, which is up for reauthorization, provides	 
transitional housing to help veterans prepare for permanent	 
housing. As requested, GAO reviewed (1) VA homeless veterans	 
estimates and the number of transitional housing beds, (2) the	 
extent of collaboration involved in the provision of GPD and	 
related services, and (3) VA's assessment of GPD program	 
performance. GAO analyzed VA data and methods used for the	 
homeless estimates and performance assessment, and visited	 
selected GPD providers in four states to observe the extent of	 
collaboration.							 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-859 					        
    ACCNO:   A60605						        
  TITLE:     Homeless Veterans Programs: Improved Communications and  
Follow-up Could Further Enhance the Grant and Per Diem Program	 
     DATE:   09/11/2006 
  SUBJECT:   Alcohol or drug abuse problems			 
	     Homelessness					 
	     Housing						 
	     Housing programs					 
	     Interagency relations				 
	     Mental illnesses					 
	     Performance appraisal				 
	     Program evaluation 				 
	     Veterans						 
	     Veterans benefits					 
	     VA Homeless Providers Grant and Per Diem		 
	     Program						 
                                                                 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-859

     

     * Results in Brief
     * Background
          * GPD Transitional Housing Program for Homeless Veterans
          * Veteran Eligibility for the GPD Program
          * Characteristics of Veterans Eligible for the GPD Program
          * Roles of Various Agencies Serving Homeless Veterans
     * VA Estimates about 194,000 Veterans Are Homeless and Has Increased Its
       Capacity to Provide Transitional Housing
          * VA Considers Its Homeless Veterans Estimate to Be the Best
            Available
          * VA Expanded GPD Program Capacity and Plans Further Expansion to
            Help Meet Homeless Veterans' Needs
     * GPD Providers Collaborate to Offer a Range of Services but Still Face
       Challenges in Helping Veterans
          * GPD Providers Create Partnerships to Help Veterans Meet Program
            Goals, but Resource Gaps Remain
          * Communication of Program Polices May Affect Providers' Ability to
            Serve Veterans
     * VA Data Show That the GPD Program Helps Veterans Get Housing and
       Income, but Data Are Limited on Veterans' Circumstances after They
       Leave the Program
          * Many Veterans Attain Stable Housing, Income, and Greater
            Self-Determination Immediately upon Leaving the Program,
            According to VA Data
               * Stability in Independent and Secured Housing
               * Income from Employment or Financial Benefits
               * Greater Self-Determination in Terms of Improved Functioning
                 in Several Areas
               * In 2006 VA Took Steps to Help Ensure That VA Liaisons
                 Conduct Required Reviews of GPD Provider Performance
          * VA Does Not Routinely Collect Data on Veterans' Long-Term
            Success, but Recent Study May Provide Insights on How Veterans
            Fare a Year after Leaving the Program
     * Conclusions
     * Recommendations for Executive Action
     * Agency Comments and Our Evaluation
     * Veterans Health Administration Programs for Homeless Veterans
          * Health Care for Homeless Veterans (HCHV) including Contracted
            Residential Treatment
          * Homeless Domiciliary Residential Rehabilitation and Treatment
            Program
          * Homeless Compensated Work Therapy/Transitional Residence
          * Loan Guarantee for Multifamily Transitional Housing
          * Housing and Urban Development-VA Supported Housing
     * Veterans Benefits Administration Programs for Homeless Veterans
          * Veterans Benefits Administration Outreach
          * Acquired Property Sales for Homeless Providers
          * Labor-VA Incarcerated Veterans' Transition Program
     * GAO Contact
     * Acknowledgments
     * GAO's Mission
     * Obtaining Copies of GAO Reports and Testimony
          * Order by Mail or Phone
     * To Report Fraud, Waste, and Abuse in Federal Programs
     * Congressional Relations
     * Public Affairs

Report to the Chairman, Committee on Veterans' Affairs, House of
Representatives

United States Government Accountability Office

GAO

September 2006

HOMELESS VETERANS PROGRAMS

Improved Communications and Follow-up Could Further Enhance the Grant and
Per Diem Program

GAO-06-859

Contents

Letter 1

Results in Brief 2
Background 4
VA Estimates about 194,000 Veterans Are Homeless and Has Increased Its
Capacity to Provide Transitional Housing 12
GPD Providers Collaborate to Offer a Range of Services but Still Face
Challenges in Helping Veterans 19
VA Data Show That the GPD Program Helps Veterans Get Housing and Income,
but Data Are Limited on Veterans' Circumstances after They Leave the
Program 26
Conclusions 34
Recommendations for Executive Action 34
Agency Comments and Our Evaluation 35
Appendix I Scope and Methodology 37
Appendix II VA's Programs for Homeless Veterans Other than the GPD Program
41
Appendix III Range of Services Offered by GPD Programs Nationwide 44
Appendix IV Participant Outcomes for the Grant and Per Diem Program 47
Appendix V Comments from the Department of Veterans Affairs 49
Appendix VI GAO Contact and Staff Acknowledgments 53
Related GAO Products 54

Tables

Table 1: Available and Needed Transitional Beds for Homeless Veterans,
Fiscal Year 2005 18
Table 2: Examples of Services and Partners That Worked with GPD Providers
We Visited 21
Table 3: Numbers and Percentages of Veterans Leaving the GPD Program with
Employment or Benefit Income, Fiscal Years 2000 through 2005 29
Table 4: Number of Veterans Leaving GPD Program and Percentage with
Specific Problems at Entry, Fiscal Years 2000 and 2005 31
Table 5: Features of GPD Programs That GAO Visited 38
Table 6: Percentage of GPD Facilities Reporting They Provided Selected
Services by Method 45
Table 7: Number Served by VA's Health Care for Homeless Veterans and Grant
and Per Diem Program and Veterans' Outcomes, Fiscal years 2000 through
2005 48

Figures

Figure 1: VA Services and Programs for Homeless Veterans 5
Figure 2: Interiors and Exteriors of Selected GPD Buildings That GAO
Toured 8
Figure 3: VA Estimates of Homeless Veterans Nationwide, Fiscal Years 2000
through 2005 14
Figure 4: Number of GPD Beds Compared to Admissions of Homeless Veterans,
Fiscal Years 2000 through 2005 16
Figure 5: Distribution of the Beds Available under the GPD Program in May
2006 17
Figure 6: Flow of Policy and Program Information from VA to GPD Providers
24
Figure 7: Percentage of Veterans with Independent or Secured Housing upon
Leaving GPD Program, Fiscal Years 2000 through 2005 28
Figure 8: Percentage of Veterans Leaving the GPD Program with Greater
Self-Determination, Fiscal Years 2000 through 2005 30

Abbreviations

CHALENG Community Homelessness Assessment, Local Education and Networking
Group for Veterans

DOL Department of Labor

GPD Homeless Providers Grant and Per Diem

HCHV Health Care for Homeless Veterans

HHS Department of Health and Human Services

HUD Department of Housing and Urban Development

NEPEC Northeast Program Evaluation Center

OIG Office of Inspector General

VA Department of Veterans Affairs

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. However, 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.

United States Government Accountability Office

Washington, DC 20548

September 11, 2006

The Honorable Steve Buyer Chairman Committee on Veterans' Affairs House of
Representatives

Dear Chairman Buyer:

On any given night in the United States, an estimated 700,000 people,
including veterans, are homeless and sleep on the streets or in shelters.
Veterans constitute about one-third of the adult homeless population,
according to the Department of Veterans Affairs (VA). Many veterans who
are not homeless may be at risk of homelessness as a result of poverty,
lack of support from family and friends, or precarious living conditions
in overcrowded or substandard housing.

To help address the needs of homeless veterans, VA operates several
programs, the largest of which is the Homeless Providers Grant and Per
Diem (GPD) program. Scheduled for reauthorization in 2007, this program
provides a transitional setting to help veterans prepare for permanent
housing. The program is not intended to serve all homeless veterans but is
focused instead on serving those who are most in need, including veterans
whose circumstances make them likely to remain homeless unless they
receive assistance, such as those who have had problems with mental
illness, substance abuse, or both. Through a network of local nonprofit or
public agencies, the program provides beds to homeless veterans in
settings free of drugs and alcohol that are supervised 24 hours a day, 7
days a week. Program rules generally allow veterans to stay with a single
GPD provider for 2 years, but providers have the flexibility to set
shorter time frames. In addition, veterans are generally limited to a
total of three stays in the program over their lifetime. The program's
goals are to help homeless veterans achieve residential stability,
increase their skill levels or income, and attain greater
self-determination.

As Congress considers the reauthorization of the GPD program, you asked us
to review (1) VA estimates of the total number of homeless veterans and
the number of transitional beds available, (2) the extent of collaboration
involved in the provision of GPD and related services, and (3) VA's
assessment of GPD program performance.

In examining VA's estimates of the number of homeless veterans, we
reviewed relevant reports and interviewed outside experts as well as
officials with the Bureau of the Census, the Department of Housing and
Urban Development (HUD), and VA's Community Homelessness Assessment, Local
Education and Networking Group for Veterans (CHALENG). To assess the
extent of coordination among community partners serving homeless veterans,
we visited 13 GPD providers located in California, Florida, Massachusetts,
and Wisconsin, including some in rural areas as well as large cities. In
addition, we analyzed data from a survey of GPD providers conducted by
VA's Northeast Program Evaluation Center (NEPEC) and attended a meeting of
VA's Advisory Committee on Homeless Veterans. We focused our review on
those GPD providers serving homeless veterans in general rather than
special subgroups, such as the chronically mentally ill. In each of these
locations, we interviewed local VA officials, GPD staff, community
partners, and, where possible, current and former program participants. To
develop information on GPD performance, we interviewed officials and
analyzed data from NEPEC and VA's national program office. Data obtained
were considered sufficiently reliable for our purposes. We coordinated
with VA's Office of Inspector General so that our review complemented but
did not duplicate its recent review related to GPD financial management
and oversight issues. 1 We conducted our work between August 2005 and July
2006 in accordance with generally accepted government auditing standards.
For more information on our scope and methodology, see appendix I.

Results in Brief

VA reports that about 194,000 veterans were homeless nationwide on a given
night in fiscal year 2005-an estimate that VA officials consider the best
available. VA changed its estimation process in 2004 to provide a snapshot
of the number of homeless veterans at a given point in time, as opposed to
an aggregate total of veterans who were homeless over the course of the
year. Earlier estimates combined these aggregate totals with the snapshot
data. While VA officials consider the current estimate to be more reliable
than those for earlier years, the agency believes the estimate to be on
the low side because some veterans cannot be located at the time the
counts are taken. To accommodate veterans ready and willing to assume the
responsibilities involved in transitional housing, VA reports that a total
of 45,000 transitional beds are needed. VA has identified 35,400 beds that
are available from various sources, including the GPD program, resulting
in a shortfall of about 9,600 beds. In fiscal year 2005, the GPD program
had about 8,000 beds available for homeless veterans. Because veterans
only stayed in GPD beds on average about 4 months, the GPD program was
able to admit over 16,000 veterans over this same period. VA officials
told us that they have plans to expand the GPD program by 2,200 beds in
the near future. As the GPD program continues to grow, VA also recognizes
that it will have to accommodate the needs of the changing homeless
veteran population, including increasing numbers of women and veterans
with dependents.

1 VA's Office of Inspector General reviewed the GPD program and planned to
issue a report in September 2006 titled Evaluation of the Veterans Health
Administration Homeless Grant and Per Diem Program that will be available
on the Internet.

2 Some medical centers continue to fund contracted residential treatment
from their own budgets. For more on earlier VA programs serving homeless
veterans, see GAO, Homeless Veterans: VA Expands Partnerships, but
Homeless Program Effectiveness Is Unclear, GAO/HEHS-99-53 (Washington,
D.C.: Apr. 1, 1999).

3Throughout this report, we use the term "GPD provider" to refer to a
locally run program. In some cases a single organization may have several
GPD grants for housing at different locations, and we generally report
this as multiple providers.

4 VA granted extensions to about 1 percent of the veterans who left the
program in fiscal year 2005. The rules allow extensions when permanent
housing for the veteran has not been located or the veteran requires
additional time to prepare for independent living.

5 In contrast to the GPD program, veterans must meet the minimum length of
service requirements of in 38 U.S.C. S:5303A in order to be eligible for
VA health care. In certain cases veterans with dishonorable discharges may
obtain an upgrade to their discharge status and thus become eligible for
the GPD program or for VA medical care.

6Veterans must constitute at least 75 percent of participants in
facilities that have received GPD capital grants.

7The definitions appear at 42 U.S.C. S: 11302 and 38 C.F.R. S: 61.1.

8To assist Continuums in conducting counts of the homeless, HUD issued A
Guide to Counting Unsheltered Homeless People, which is available on the
Internet.

9For the committee's recommendations and VA's responses, see Department of
Veterans Affairs, 2005 Annual Report of the Advisory Committee on Homeless
Veterans: Reaching Out to Homeless Veterans (Washington, D.C.: July 2005).

10The chronic homeless are unaccompanied individuals with disabling
conditions who have either been continuously homeless for a year or have
had at least four episodes of homelessness in the past 3 years. An
estimated 63,000 veterans were considered chronically homeless in 2005.

11For the fiscal year 2005 report, see VA, Community Homelessness
Assessment, Local Education and Networking Group (CHALENG) for Veterans:
The Twelfth Annual Progress Report on Public Law105-114, Services for
Homeless Veterans Assessment and Coordination, (Washington, D.C.: Apr. 15,
2006).

12 The estimate of homeless veterans is derived from the CHALENG survey of
designated local VA officials who are asked to provide the highest number
of homeless veterans estimated in their service area on one day of the
official's choosing in fiscal year 2005.

13The Bureau of the Census has had difficulty enumerating the overall
homeless population, as we reported in GAO, Decennial Census: Methods for
Collecting and Reporting Data on the Homeless and Others without
Conventional Housing Need Refinement, GAO-03-227 (Washington, D.C.: Jan.
17, 2003). A Census official we interviewed cautioned that the 2010 Census
may not enumerate homeless veterans.

14 VA issued a directive for a onetime dental care opportunity for
homeless veterans (VHA Directive 2002-080) in line with 38 U.S.C. S: 101
note. VA officials told us that funding was provided in 2006 to implement
this directive.

15 For more information on DOL programs, see GAO, Homeless Veterans: Job
Retention Goal Under Development for DOL's Homeless Veterans'
Reintegration Program, GAO-05-654T (Washington, D.C.: May 4, 2005).

16Through the Continuum of Care, HUD contracts with public housing
agencies for the rehabilitation of residential properties that provide
multiple single room dwelling units. These agencies make Section 8 rental
assistance payments generally covering the difference between a portion of
the tenant's income (normally 30 percent) and the unit's rent to
participating owners (i.e., landlords) on behalf of homeless individuals
who rent the rehabilitated dwellings.

17 According to VA, in fiscal years 2005 and 2006 it had allocated funding
for a total of 97 full-time liaisons. As of the time of our review, some
sites were still going through the recruitment and hiring processes to
fill these positions.

18 VA may waive the episode requirement if the services offered are
different from those previously provided and may lead to a successful
outcome. The VA liaisons must review and approve or deny the waiver based
on their best clinical assessment of the individual case.

19Since fiscal year 2002, VA's strategic plan has included a performance
target to capture the housing status of veterans discharged from three of
its transitional housing programs, including the GPD program. VA has
gradually increased its target from 65 percent in fiscal year 2002 to 79
percent in fiscal year 2005. VA estimates that it exceeded this target in
fiscal year 2005.

20 VA also asks participants for their evaluations after they have been in
the program for 1 month. Nearly half of the participants completed the
surveys in fiscal year 2005. Most reported satisfaction with the GPD,
rating it at 3.2 on a scale where 4 is the highest possible score, and
with their VA case managers, rating them at 4.6 on a scale where 6 is the
highest possible score.

21 This effort has been possible, according to VA, in part because
increased funds have made it possible for more liaisons to work with the
GPD program on a full-time rather than a part-time basis.

22VA has also conducted other follow-up studies designed to test
innovative approaches to serving homeless veterans, including ways to
improve employment outcomes, ensure the safety and serve the needs of
female veterans, and intervene on behalf of veterans dually diagnosed with
both mental health and substance abuse problems.

The GPD providers that we visited often collaborated with public and
nonprofit agencies in helping veterans to recover from substance abuse or
mental illness and obtain permanent housing, employment, financial
stability, and services needed to enhance their ability to live
independently. While GPD providers were generally able to build successful
partnerships, most of them identified resource and communications gaps
that presented challenges to delivering certain services. For example,
providers reported difficulties in locating affordable permanent housing
for veterans ready to leave the program because of shortages in their
communities. In addition we found that those responsible for program
implementation did not always understand the policies. Some GPD providers
believed that homeless veterans were eligible for the GPD program only if
they were eligible for VA health care. This assumption was incorrect and
may have had the effect of erroneously turning away veterans seeking to
enter the GPD program. There were also instances in which GPD providers
did not understand that veterans may be able to exceed the 3-stay lifetime
limit under certain conditions. This assumption, also incorrect, could
keep veterans from obtaining needed care.

VA assesses performance in two ways-the veterans' circumstances at the
time they leave the program and the ability of individual GPD providers to
meet their own objectives-but VA generally does not know how veterans are
faring months or years later. When veterans leave the program for any
reason, VA collects information on their immediate success in obtaining
housing, income, and greater self-determination-the primary measures of
overall GPD program performance. VA reports that of all veterans leaving
the program in fiscal year 2005, half had successfully arranged
independent housing, one-third had jobs, over one-third were receiving
public benefits, and 57 to 69 percent showed progress with substance
abuse, mental health or medical problems or demonstrated greater
self-determination in other ways. In addition, in 2006 VA took steps to
ensure that its local staff conduct annual reviews to determine if the GPD
providers are meeting their objectives. VA does not require providers to
collect data from veterans months or years after they leave the program,
although many providers attempt to maintain contact with former
participants. Some indication of how veterans are faring after they leave
the program should be available from VA's recent follow-up study of 520
program participants. Preliminary results of this study indicate that
veterans maintained positive housing outcomes 1 year after leaving the GPD
program.

To further strengthen VA's ability to help homeless veterans, we are
recommending that VA take steps to ensure that GPD policies and procedures
are consistently understood and to explore feasible means of obtaining
information about the circumstances of veterans after they leave the GPD
program. In its comments on a draft of this report, VA concurred with our
recommendations and described several initiatives planned or under way to
address some issues raised in our report as well as other challenges the
GPD program faces.

Background

The GPD program is one of nine VA programs that specialize in serving
homeless veterans. Six of these programs fall under the responsibility of
the Veterans Health Administration, which obligated about $224 million in
fiscal year 2006 for these programs as well as $1.2 billion for outreach
and treatment of homeless veterans. Outreach is considered particularly
important to locate and serve veterans living on the street and in
temporary shelters who otherwise would not seek assistance. Treatment
involves primary and specialty medical care, mental health care, and
alcohol and drug abuse services for eligible homeless veterans. Three of
the nine programs are run jointly or solely by the Veterans Benefits
Administration that also serves homeless veterans as part of its broader
mission to provide disability compensation and pensions to eligible
veterans. Figure 1 illustrates some of the key programs and services for
homeless veterans-including the GPD program that is the focus of this
report-provided by VA. (App. II provides a general description of the
eight programs not otherwise covered in this report.)

Figure 1: VA Services and Programs for Homeless Veterans

aThis program is a joint initiative with VHA.

bHUD provides the housing subsidy; VA provides case management services.

GPD Transitional Housing Program for Homeless Veterans

The GPD program--VA's major transitional housing program for homeless
veterans--spent about $67 million in fiscal year 2005. It became VA's
largest program for homeless veterans after fiscal year 2002, when VA
began to increase GPD program capacity and phase out national funding for
the more costly contracted residential treatment-another of VA's
transitional housing programs. 2 To operate the GPD program at the local
level, nonprofit and public agencies compete for grants. The program
provides two basic types of grants-capital grants to pay for the buildings
that house homeless veterans and per diem grants for the day-to-day
operational expenses.

           o  Capital grants cover up to 65 percent of housing acquisition,
           construction, or renovation costs and require that agencies
           receiving the grants cover the remaining costs through other
           funding sources. Generally, agencies that have received capital
           grants are considered for subsequent per diem grants, so that the
           VA investment can be realized and the buildings can provide
           operational beds.

           o  Per diem grants support the operations of about 300 GPD
           providers 3 nationwide. The per diem grants pay a fixed dollar
           amount for each day an authorized bed is occupied by an eligible
           veteran up to the maximum number of beds allowed by the grant.
           Generally under this grant, VA does not pay for empty beds. VA
           makes payments after an agency has housed the veteran, on a cost
           reimbursement basis, and the agency may use the payments to offset
           operating costs, such as staff salaries and utilities. By law, the
           per diem reimbursement cannot exceed a fixed rate, which was
           $29.31 per person per day in 2006. Reimbursement may be lower for
           providers receiving funds for the same purpose from other sources.

On a limited basis, special needs grants are available to cover the
additional costs of serving women, frail elderly, terminally ill, or
chronically mentally ill veterans. Although the primary focus of the GPD
program is housing, grants may also be used for transport or to operate
daytime service centers that do not provide overnight accommodations.
According to VA, in fiscal year 2005, GPD grants supported about 75 vans
that were used to conduct outreach and transport homeless veterans to
medical and other appointments. Also, 23 service centers were operating
with GPD support.

Most GPD providers have 50 or fewer beds available for homeless veterans,
with the majority of providers having 25 or fewer. Accommodations vary and
may range from rooms in multistory buildings in the inner city to rooms in
detached homes in suburban residential neighborhoods. Veterans may sleep
in barracks-style bunk beds in a room shared by several other participants
or may have their own rooms. Figure 2 shows the exteriors and interiors of
selected GPD buildings we visited.

Figure 2: Interiors and Exteriors of Selected GPD Buildings That GAO
Toured

Generally housing is either male only or has separate sleeping areas for
males and females. Multipurpose rooms may be available for television,
games, and conversation, as well as communal kitchen facilities where
meals can be purchased or made by the participants themselves. Not all GPD
providers supply food. Some may assist the participants in obtaining items
from community food banks. GPD providers may require veterans to pay rent,
but the rent cannot exceed 30 percent of a veteran's income, after
deducting the costs of medical, child care, and court-ordered payments. In
addition, veterans may be charged fees for other services not supported by
the GPD grant, such as cable television. According to VA rules, veterans
may stay with a single GPD provider for 24 months or longer under certain
conditions. 4 GPD providers may specify shorter limits such as 3, 6, or 12
months. In fiscal year 2005, the average stay for veterans was about 4
months with a single GPD provider.

Veteran Eligibility for the GPD Program

To meet VA's minimum eligibility requirements for the program, individuals
must be veterans and must be homeless. A veteran is defined as an
individual who has been discharged or released from active military
service and includes members of the Reserves and National Guard with
active federal service. Although the GPD program definition excludes
individuals who have received a dishonorable discharge, it is less
restrictive in terms of length of service requirements. As a result, some
homeless veterans may be eligible for the GPD program and not eligible for
VA health care. 5 VA does not pay for spouses and children of veterans who
are not themselves veterans, but they may be served by GPD providers using
other funds. 6 Consistent with the definition used in many other federal
programs, VA defines a homeless individual as a person who lacks a fixed,
regular, adequate nighttime residence and instead stays at night in a
shelter, institution, or public or private place not designed for regular
sleeping accommodations. 7 Prison inmates are not deemed homeless, but may
be at risk of homelessness and may be eligible for the program upon their
release. GPD providers determine if potential participants are homeless,
but VA officials determine if potential participants meet the program's
definition of veteran. VA officials are also responsible for determining
whether veterans have exceeded their lifetime limit of three stays in a
GPD program and for issuing a waiver to that rule when appropriate.

Prospective GPD providers may identify additional eligibility requirements
in their grant documents. Because the providers are responsible for
providing a clean and sober environment that is free of illicit drugs,
about two-thirds of providers require that veterans entering the program
be sober and free from alcohol and drug use for a given length of time.
The time frames set by many providers range from 1 to 30 days of sobriety.
Many providers also conduct drug tests of veterans after they enter the
program to ensure their continued sobriety. Most providers will not accept
veterans considered to be a danger to themselves or others, in need of
detoxification, or under the influence of drugs or alcohol. About
one-fifth of providers also exclude veterans who are considered seriously
mentally ill, because the providers may not be able to provide adequate
care.

Characteristics of Veterans Eligible for the GPD Program

The GPD program is focused primarily on helping those most in
need-veterans who might remain homeless for long periods of time if no
intervention occurs-and is not intended to serve all homeless veterans.
About two-thirds of homeless veterans in the program in fiscal year 2005
had struggled with alcohol, drug, medical, or mental health problems.
About 40 percent of homeless veterans seen by VA had served during the
Vietnam era, and most of the remaining homeless veterans served after that
war, including over 2,500 who served in military operations in the Persian
Gulf, Afghanistan, and Iraq. Almost all homeless veterans seen by VA are
males; about half are between 45 and 54 years old, one-quarter are older,
and one-quarter are younger. African-Americans are disproportionately
represented, constituting the largest racial group at 47 percent; whites
are the next largest group at 45 percent. About 75 percent of veterans are
either divorced or never married.

Roles of Various Agencies Serving Homeless Veterans

The complex problems faced by homeless veterans require a system of
comprehensive, integrated services that often involves multiple
organizations. Key federal agencies with programs specifically targeted to
the homeless, including veterans, are HUD, the Department of Health and
Human Services (HHS), and the Department of Labor (DOL). HUD makes funds
available to bring together community organizations to plan and coordinate
service delivery through local or regional networks designated as the
"Continuums of Care." In their planning role, the Continuums arrange for
counts of the homeless in their area, and since 2003, are required to
report the number for a given point in time and to do so at least every 2
years. 8 Further, as part of their coordination role, the Continuums
review agency applications for certain HUD grants. HUD also funds
emergency shelters that are open seasonally or year-round for temporary,
overnight accommodations. In addition, HUD is the only federal agency that
is authorized to provide permanent subsidized housing for the homeless.
HHS specializes in funding health care and researching the needs of
homeless with substance abuse and mental health issues. DOL, like VA, has
programs targeted specifically to veterans within the homeless population,
with DOL's emphasis on helping veterans obtain employment. Charities,
businesses, and state and local governments are also involved in meeting
the needs of homeless veterans and, in some cases, providing funding to
GPD providers.

At the federal level, VA works with these and other federal agencies
through two key committees. VA's Advisory Committee on Homeless Veterans
is responsible for assessing the needs of homeless veterans and
determining if VA and others are meeting these needs. The committee
comprises homeless veterans, experts and advocates, community-based
service providers, state and federal government officials, and
representatives of veterans' service organizations. The committee has made
several recommendations on improvements to homeless veterans' programs,
including the GPD program, some of which have been implemented. In 2004
the committee urged VA to fund GPD providers serving veterans with special
needs, especially female veterans; in fiscal year 2005 there were 29
programs of this kind, including 8 for female veterans. 9

VA is also a participant on the Interagency Council on Homelessness, which
coordinates the federal response to homelessness and works with state and
local governments to develop plans for ending chronic homelessness among
individuals, including veterans, in 10 years. 10 Although the chronic
homeless represent only 10 to 20 percent of all homeless adults, they take
up roughly half of all shelter beds and also use a disproportionate share
of resources for the homeless.

At the local level, VA works with various agencies through the Community
Homelessness Assessment, Local Education and Networking Groups for
Veterans, referred to as Project CHALENG. An arrangement of this kind is
needed, according to VA, because no single agency can provide the full
range of services required to help homeless veterans become more
productive members of society. Through CHALENG, a designated VA official
in each medical center, usually VA's homeless coordinator, reaches out to
community agencies that provide services to the homeless to raise
awareness of homeless veterans' particular needs and to plan to meet those
needs. Specific needs to be addressed include outreach, counseling, health
care, education and training, employment, and housing. Every year these VA
officials prepare estimates of the total number of homeless veterans in
their area, based on input from various sources. In addition, the
officials meet with community representatives to complete a survey of
available resources, additional resources needed, priorities for service,
and an action plan. 11

VA Estimates about 194,000 Veterans Are Homeless and Has Increased Its
Capacity to Provide Transitional Housing

VA estimates that on a given night in fiscal year 2005 about 194,000
veterans were homeless. 12 The estimate, generally lower than the numbers
reported prior to 2004, is considered by VA officials to be the best
estimate available. VA officials believe that a new methodology and use of
local HUD data has improved the estimate, although some homeless veterans
may not have been included because they could not be found when the
estimate was developed. While VA has increased its capacity to provide
transitional housing for homeless veterans in recent years, its program
planning efforts indicate that an additional 9,600 transitional housing
beds from various sources are needed to meet current demand. VA officials
report that they are working to operationalize an additional 2,200 beds
for the GPD program.

VA Considers Its Homeless Veterans Estimate to Be the Best Available

VA bases its national estimate of homeless veterans on the summation of
local estimates developed by VA officials for the areas served by VA
medical facilities. This process is part of the annual CHALENG planning
effort, which involved 135 local VA officials in 2005. Local VA officials
are not responsible for conducting their own counts of homeless veterans,
but are expected to rely on data from other groups that have collected
these data. More than 75 percent of VA officials use multiple data
sources, in part because the areas covered by VA medical facilities often
comprise several cities, counties, or even states, while local data
sources may cover one or more of these jurisdictions, but rarely cover the
full area served by the medical facility. Most often, local VA officials
rely on data collected by the HUD-funded Continuums of Care, local
governments, university researchers, or other groups along with
information from local homeless providers. The estimates reported by local
VA officials are compared to the previous year's and if they have
significantly changed, the local VA officials are asked to explain the
differences before their estimates are incorporated into the national
figure.

Prior to 2004, local VA officials used a methodology to develop their
estimates that was the equivalent of mixing apples with oranges and, as a
result, yielded less consistent, reliable counts of the homeless veteran
population. This mixed methodology combined cumulative numbers such as the
total who were homeless over the course of a year with point-in-time
numbers involving the number homeless on any given day or night. The
numbers were not comparable because over the course of a year some
individuals who were not homeless when the counts were conducted later
became homeless. Generally, the number of veterans who are homeless
sometime over the course of a year is larger than the number who are
homeless on any given night. Since 2004, local VA officials have been
directed to use point-in-time data exclusively in developing their
estimates to reflect the number of homeless veterans on any given day of
the year. VA reports that this standardized method yields more reliable
estimates than were developed for earlier years, although there may be
some veterans who cannot be located. Figure 3 shows VA's estimates of the
homeless veteran population from fiscal years 2000-2005.

Figure 3: VA Estimates of Homeless Veterans Nationwide, Fiscal Years 2000
through 2005

Recent estimates are also likely to be more reliable, according to VA,
because local VA officials increasingly use homeless data from counts
funded by HUD's Continuum of Care, which are believed to be more accurate.
In 2005, more than twice as many local VA officials used HUD counts as was
the case in 2003. HUD-funded counts in many communities are gradually
improving as the census takers increasingly seek out the "hidden" homeless
who do not contact service providers as well as the homeless who
congregate at soup kitchens and shelters. In both Atlanta and Los Angeles,
homeless individuals were hired in 2005 to assist the census takers in
locating areas where homeless individuals could be found. As a result, the
local counts that were conducted in these two communities were more
accurate than the counts conducted in earlier years, according to VA
officials.

Although VA officials believe that the number is likely an underestimate,
VA officials consider their 2005 year estimate of 194,000 homeless
veterans on any given night to be the best available. Counting the
homeless is a challenge for several reasons, as VA and other agencies have
acknowledged, 13 since the homeless are hard to locate and some may not be
included in the current estimate. Also, the number may change in relation
to social and economic factors, such as job layoffs or a tighter housing
market. In addition, veterans who are doubled up and sharing crowded
living quarters with others are considered at risk of becoming homeless
but are not included in the counts because they do not meet VA's
definition of homeless.

VA Expanded GPD Program Capacity and Plans Further Expansion to Help Meet
Homeless Veterans' Needs

Since fiscal year 2000, VA has almost quadrupled the number of available
beds and the number of admissions of homeless veterans to the GPD program
in order to address some of the needs identified through the CHALENG
survey. In fiscal year 2005, VA had the capacity to house about 8,000
veterans on any given night. However, over the course of the year, because
some veterans completed the program in a matter of months and others left
before completion, VA was able to admit about 16,600 veterans into the
program. Figure 4 illustrates the growth in GPD program capacity from
fiscal years 2000 through 2005.

Figure 4: Number of GPD Beds Compared to Admissions of Homeless Veterans,
Fiscal Years 2000 through 2005

Note: Not all beds shown were in operation for the full year; for example,
only 7,800 beds were in operation at the end of fiscal year 2005.

VA has pursued a policy of making GPD beds available in all states and the
District of Columbia, in line with the recommendation made by the VA
Advisory Committee on Homeless Veterans. As shown in figure 5, all but
three states had beds available in May 2006, and VA officials told us that
they were working with potential providers to develop the capacity in
these states. The greatest number of beds is in California (1,867 beds);
Florida and Massachusetts (430 and 378 beds, respectively); and New York,
Ohio and Pennsylvania (274, 261, and 332 beds respectively).

Figure 5: Distribution of the Beds Available under the GPD Program in May
2006

Note: VA reports that grants have been awarded to providers in Alaska,
Maine, and North Dakota to develop 20, 18 and 48 beds respectively in
those states.

VA's CHALENG report found that about 45,000 transitional housing beds were
needed in fiscal year 2005 to help homeless veterans become more socially
and economically independent. As shown in table 1, the report identified
over 35,000 transitional housing beds that were available through various
sources for this purpose--including the GPD beds, another 2,400 beds
funded by VA through its other specialized homeless programs, and
additional beds funded by other sources. Still needed were about 9,600
more transitional housing beds nationwide beyond the number currently
available to meet the demand in fiscal year 2005. To begin to address the
demand, VA officials told us that, as of May 2006, they have negotiated an
additional 2,200 beds for the GPD program that are expected to be
available in the near future.

Table 1: Available and Needed Transitional Beds for Homeless Veterans,
Fiscal Year 2005

Transitional beds needed            45,000 
VA transitional beds available      10,400 
                                              
      o  GPD program                  (8,000) 
      o  Non-GPD programsa                    
                                      (2,400) 
Other transitional beds available          
                                       25,100 
Total transitional beds availableb  35,400 
Additional beds still neededb        9,600 

Source: GAO analysis of VA data.

aBeds for VA's contracted residential treatment are not included, but VA
officials estimate about 304 beds are available.

bNumbers are CHALENG estimates rounded to nearest 100; subtotals included
in these numbers may not add to numbers shown due to rounding.

Although VA reports the need for transitional housing beds is greater than
the capacity, the demand varies throughout the year and by location. Some
GPD programs we visited had vacancies and others had waiting lists at the
time of our visit. GPD providers and VA officials identified several
reasons that beds may go unfilled at any given time. Some beds are held
for veterans who are receiving medical treatment, while others may be
unfilled as a result of the normal transition when one veteran has left
the program and another veteran will soon be entering the program.

VA officials and GPD providers also told us they expect a change in the
demographics of homeless veterans that may require them to reconsider the
type of housing and services that they are providing with GPD funds.
Specifically, VA officials expect to see more homeless women veterans and
more veterans with dependents who are in need of transitional housing. GPD
providers told us that women veterans have sought transitional housing;
some recent admissions had dependents; and a few of their beds were
occupied by the children of veterans, for whom VA could not provide
reimbursement. To meet the needs of homeless women veterans, VA has
provided additional funding in the form of special needs grants to a few
GPD programs.

GPD Providers Collaborate to Offer a Range of Services but Still Face
Challenges in Helping Veterans

GPD providers often worked with public and nonprofit agencies to offer a
spectrum of services that may help veterans meet individual and GPD
program goals. While GPD providers were generally able to build successful
partnerships, most of them identified resource gaps that presented
challenges to helping veterans, particularly affordable permanent housing.
We also found that communication issues related to program policies could
prevent veterans from being offered care. Providers did not always
understand eligibility requirements such as which veterans may be eligible
for the program and the allowable number and length of program stays.
Further, providers were not always aware of policy changes.

GPD Providers Create Partnerships to Help Veterans Meet Program Goals, but
Resource Gaps Remain

GPD providers generally created partnerships to help prepare veterans to
obtain permanent housing and, ultimately, to live independently. VA's
grant process encourages such collaboration by awarding points to GPD
program applicants that demonstrate they have relationships with other
organizations. GPD providers are to identify how they will provide
services to meet the program's goals-residential stability, increased
skill level or income, and greater self-determination. For example,
providers may identify services such as substance abuse and mental health
treatment, financial counseling, employment assistance and training,
transportation to appointments and job interviews, and related services.
We found variation in the agencies that provided these services. According
to a VA survey, most GPD providers used their own on-site staff to offer
services like case management and transportation assistance. In contrast,
mental health assessments were mostly handled indirectly, with 79 percent
of the GPD providers using the staff of other agencies, often the VA.
(More information from the survey can be found in app. III.)

The GPD providers that we visited established partnerships with state and
local government agencies, other federal agencies, and local community
organizations. Further, several of the providers that we visited
participated in the local Continuum of Care funded by HUD or in other
community coalitions, taking advantage of community networks that serve
homeless individuals. While most providers offered a range of services,
not all veterans received each service. To identify the specific services
a veteran may need, providers typically worked with veterans to develop
individual treatment plans that identified the veteran's needs on entering
the program. Table 2 lists examples of services and partners of GPD
providers we visited.

Table 2: Examples of Services and Partners That Worked with GPD Providers
We Visited

Veterans'   Select services  Partners that     
needs                        provided          
                                servicesa         
Case management and             o  VA liaison  
individual treatment plan       with GPD       
                                   provider       
Health Care Mental health       o  VA          
               treatment           o  Local area  
                                   hospitals      
                                   o  Local       
                                   organizations  
                                Substance abuse      o  VA                    
                                treatment            o  Local area hospitals  
                                                     o  Local organizations   
                                Counseling           o  VA                    
                                (family,             o  Local organizations   
                                nutritional,      
                                etc.)             
                                Medical services     o  VA                    
                                                     o  Local area hospitals  
Employment  Financial           o  Local       
and Income  counseling          organizations  
                                Employment           o  Department of Labor   
                                assistance and                                
                                training                     o  Disabled      
                                                             Veterans'        
                                                             Outreach         
                                                             Programb         
                                                             o  Homeless      
                                                             Veterans         
                                                             Reintegration    
                                                             Programc         
                                                                              
                                                     o  VA                    
                                                                              
                                                             o  Compensative  
                                                             Work Therapy     
                                                             o  Incentive     
                                                             Therapyd         
                                                                              
                                                     o  State and local       
                                                     training programs        
                                                     o  Local organizations   
                                                     and colleges             
                                Assistance with      o  VA                    
                                getting benefits     o  Social Security       
                                                     Administration           
                                                     representative           
                                                     o  State/county benefits 
                                                     counselors               
                                                     o  Veterans service      
                                                     organizations            
After       Stable housing      o  State and   
leaving GPD                     local programs 
program                         o  HUD         
                                Follow-up care       o  VA                    
                                and supportive       o  Local organizations   
                                servicese         
Other needs Legal assistance    o  Local       
                                   organizations  
                                   and law        
                                   offices        
                                   o  Local       
                                   colleges       
                                   o  Outreach to 
                                   local jails    
               Transportation      o  VA GPD van  
                                   grants         
                                   o              
                                   Relationship   
                                   with local     
                                   transit        
                                   authority      

Source: GAO analysis of GPD provider partnerships.

aGPD provider staff also may have been directly involved in providing
services in any of these partnership examples.

bProgram provides funding through state employment security agencies to
support dedicated staff positions to develop and provide employment and
job training opportunities for disabled and other qualified veterans.

cProgram provides services to assist in reintegrating homeless veterans
into meaningful employment within the labor force.

dProgram helps veterans regain work habits and skills by participating in
various work situations within VA as part of their treatment or
rehabilitative programs.

eSupportive services for veterans who leave the GPD program may include
health care services rendered during a veteran's GPD program stay, as well
as other services to help veterans maintain housing.

GPD programs often collaborated with VA and others to provide health
care-related services-such as mental health and substance abuse treatment,
and family and nutritional counseling-to help veterans become more
self-sufficient in their day-to-day activities. Several programs hosted
Alcoholics Anonymous meetings and other counseling services, while some
GPD programs expected veterans to attend regular meetings elsewhere in the
community. At least two GPD providers we visited provided their own
substance abuse treatment and did not rely on community partners to
provide such services. At least two other providers that referred veterans
to VA for substance abuse treatment expressed concerns about waiting lists
for that service, making it hard for veterans to access care immediately.
Typically, a VA local medical center provided veterans with primary and
specialized health care. However, GPD providers sometimes expressed
concerns about difficulties obtaining dental care. 14 To meet the needs of
veterans who were not eligible for VA health care, GPD providers made
other arrangements. For example, a program in the Boston area partnered
with the local hospital which provided free health care to homeless
veterans who were in the GPD program but were ineligible for VA health
care. We also found that many providers either used their own staff or
used partners' staff to provide mental health services and family and
nutritional counseling services.

All providers we visited tried to help veterans obtain financial benefits
or employment. Some had staff who assessed a veteran's potential
eligibility for public benefits such as food stamps, Supplemental Security
Income, or Social Security Disability Insurance. Other providers relied on
relationships with local or state officials to provide this assessment.
For example, a Wisconsin GPD provider worked with a county veterans'
service officer who reviewed veterans' eligibility for state and federal
benefits. The provider also had a relationship with a county employment
representative who came to the GPD facility to discuss job searches,
training, and other employment issues with veterans. Several providers
were receiving DOL grants to provide employment training services, worked
with local colleges, or relied on other local programs to help veterans to
increase skills. 15 However, a lack of available jobs in an area may
sometimes pose problems to finding employment for veterans.

Most of the GPD providers in the areas that we visited worked with
community partners to obtain permanent housing for veterans ready to leave
the GPD program, but indicated this was sometimes difficult because of
limited affordable permanent housing. Some providers had established
extensive partnerships with organizations that provide or find affordable
permanent housing. For instance, several of the providers worked with the
local HUD-funded Continuum of Care network to identify permanent housing
resources. Some providers had or were applying for HUD funds to build
single room occupancy housing units that could serve as a transition to
more permanent long-term housing. 16 As at least one provider mentioned,
veterans sometimes become resourceful and agree to share apartments. In
some instances, providers have asked for an extension to allow veterans to
stay until housing becomes available.

GPD providers and VA staff coordinated with community resources to help
address other issues that they identified that might also present
obstacles for transitioning veterans out of homelessness. For example,
staff in some locations indicated that such legal issues as criminal
records or credit problems may preclude veterans from obtaining employment
and housing. To help overcome these issues, some GPD providers worked with
lawyers who provided services at no cost or other volunteer organizations.
Staff in some of the locations also reported that transportation issues
made it difficult for veterans to get to medical appointments or
employment-related activities. To help address potential transportation
difficulties, some providers received GPD grants to purchase vans. One
provider that we visited partnered with the local transit company that
provided subsidies to homeless veterans. This option is not always
available, however, and transportation remained an issue in areas not near
a medical center.

Communication of Program Polices May Affect Providers' Ability to Serve
Veterans

VA has five staff in the national program office who administer the GPD
program through a network of 21 regional homeless coordinators and 136
local VA liaisons. While program policies are developed at the national
level by the GPD program staff, the local VA liaisons designated by VA
medical centers have primary responsibility for communicating with GPD
providers in their area. Figure 6 depicts the flow of information about
the GPD program.

Figure 6: Flow of Policy and Program Information from VA to GPD Providers

The VA liaisons may serve in a full-time or part-time capacity, in part
depending on the number of GPD beds in the area served by the VA medical
centers and the number of admissions per year. In fiscal year 2006, there
were 60 full-time liaisons and another 76 individuals serving as part-time
liaisons in addition to their other VA duties. Liaisons sometimes found it
hard to readily assist providers, according to some staff we met, because
of the liaisons' large caseloads and multiple GPD
responsibilities-including eligibility determination, verification of
intake and discharge information, case management, fiscal oversight,
monitoring program compliance and inspections of GPD facilities, among
other duties. To help address this issue, VA has set aside additional
funding for more full-time liaisons. 17

The program office communicates with GPD providers and VA liaisons through
written guidance and teleconferences. VA provides liaisons with a
guidebook about their responsibilities and the program rules as well as a
manual prepared by NEPEC on the forms to be completed for all program
participants. To stay up-to-date on GPD program policies, liaisons
participated in monthly conference calls and also had the opportunity to
attend a conference conducted by the GPD program office in 2004. The
program office recently held a training seminar for new liaisons and also
offers training via phone. VA also gives GPD providers program handbooks
and holds monthly conference calls to discuss program rules. In addition,
some of the VA medical centers we visited held meetings with local GPD
program providers in their areas to share information.

Despite VA's efforts, we found that some providers did not understand all
of the GPD program policies. Some misunderstandings could affect a
veteran's ability to get-and a GPD provider's ability to offer-care. For
instance, two providers said that VA staff told them that veterans
eligible to participate in the GPD program were also required to be
eligible for VA health care, but this is not the case. Similarly, in
another location, the local VA liaison and a provider both told us that
they had received information from the GPD program office indicating that
the total lifetime length of stay was 2 years, but the GPD program
officials told us this interpretation of the information that they
provided is incorrect. Elsewhere several providers understood the lifetime
limit of three GPD stays but may not have known or believed that waivers
to this rule could be granted. They argued that the limit could hinder a
veteran's ability to participate in the GPD program if participation
involved phased care offered by separate GPD providers, each specializing
in certain phases of treatment, such as detoxification or job preparation.
Since each phase of treatment is counted as one GPD stay, veterans may
exhaust their 3-stay limit before they have received services vital to
their improved functioning. Although VA has the authority to waive the
3-stay limit in such cases, these providers did not seem to understand
that this option was available to them. 18 In addition, providers were not
always aware of changes in the GPD program in a timely fashion; sometimes
not at all. For example, not all GPD providers knew in 2006 that their
program's inspections would include a review of whether they were meeting
the objectives described in their GPD grant documents.

VA recognizes that communication to providers and liaisons needs to be
improved. In its fiscal year 2005 report, the VA Advisory Committee on
Homeless Veterans recommended that VA hold an annual conference and that
each GPD provider have an opportunity to attend at least one such
conference. The purpose of the conference would be to improve
communications, program compliance, and treatment strategies. In the
spring of 2006 when the committee reconvened, VA had not yet accepted the
committee's recommendation.

VA Data Show That the GPD Program Helps Veterans Get Housing and Income,
but Data Are Limited on Veterans' Circumstances after They Leave the
Program

VA data show that in fiscal years 2000-2005 a steady or increasing
percentage of veterans had stable housing, income, and greater
self-determination at the time they left the GPD program. These national
performance results are derived from standard forms filled out by VA staff
or by provider staff with VA's review and sign-off for every veteran who
leaves the program for any reason. While the veterans' success is VA's
primary measure of program performance, in 2006 VA took steps to ensure
that the performance of individual GPD providers would also be reviewed,
in line with a recommendation of VA's Office of Inspector General (OIG).
Some GPD providers we visited had stated in their grant documents that a
certain percentage of veterans they served would have permanent housing or
employment a year after they left the program. Also, VA recently completed
a onetime study looking at longer-term outcomes for homeless veterans,
including 520 who participated in the GPD program, and preliminary results
show that positive housing outcomes were maintained 1 year after veterans
left the GPD program. However, VA does not routinely collect follow-up
information to determine the status of participants at specified times
after they leave the program and may not be able to rely on the results of
its study to determine the success of future program participants.

Many Veterans Attain Stable Housing, Income, and Greater
Self-Determination Immediately upon Leaving the Program, According to VA
Data

The following sections compare VA's GPD performance data from fiscal year
2005 with data from fiscal years 2000 through 2004.

Stability in Independent and Secured Housing

VA reports that about 81 percent of veterans had arranged some form of
housing at the time they left the GPD program in fiscal year 2005, a
significant improvement over the 56 percent with housing in fiscal year
2000. VA considers the program successful if veterans have obtained either
independent or secured housing. 19 Independent housing comprises
apartments, rooms, or houses, while secured housing includes transitional
housing programs, halfway houses, hospitals, nursing homes, or similar
facilities. Most of the improvement in housing outcomes has occurred in
independent housing. While independent housing may be a more desirable
outcome, for some veterans, including those with severe disabilities,
secured housing may be more appropriate. Figure 7 shows the percentages of
veterans who had arranged housing when they left the GPD program in fiscal
years 2000 through 2005.

Figure 7: Percentage of Veterans with Independent or Secured Housing upon
Leaving GPD Program, Fiscal Years 2000 through 2005

In its annual reports, VA compares the housing arrangements of veterans
who successfully met provider requirements with those who did not. As
might be expected, proportionately more veterans who met requirements had
obtained independent housing in fiscal year 2005--nearly 70
percent-compared to the 40 percent with independent housing who had not
met provider requirements. In terms of numbers, about half of the 15,000
veterans who left the program in fiscal year 2005 were considered by the
GPD providers to have met program requirements, an improvement over
earlier years. Of the approximately 7,500 veterans remaining, about half
dropped out and the other half violated program rules, such as rules on
maintaining sobriety, or they left for other reasons. VA derives this
information from discharge forms completed by VA or GPD staff for all
veterans at the time they leave the program. VA's evaluation center NEPEC
aggregates this data and prepares annual reports on overall GPD program
performance. For more on this process, see appendix IV.

Income from Employment or Financial Benefits

The program goal of increased income can be achieved through maintaining
or obtaining employment or financial benefits such as VA disability
compensation or pensions, Supplemental Security Income, or food stamps.
From fiscal years 2000 to 2005, about one-third of veterans had jobs,
mostly on a full-time basis, when they left the GPD program. The number of
veterans with jobs more than tripled over the period, with about 4,900
employed in fiscal year 2005 at the time they left the program. The number
of veterans receiving VA benefits when they left the GPD program was about
3,800, while another 2,200 veterans had applied or planned to apply for VA
benefits. Table 3 shows the percentages and numbers of those employed or
receiving benefits for fiscal years 2000 through 2005, but VA did not have
data on receipt of benefits until 2003.

Table 3: Numbers and Percentages of Veterans Leaving the GPD Program with
Employment or Benefit Income, Fiscal Years 2000 through 2005

Number and             Fiscal   Fiscal   Fiscal   Fiscal   Fiscal   Fiscal 
percentage of       year 2000     year     year     year     year     year 
discharges from GPD               2001     2002     2003     2004     2005 
program with                                                      
      o  Total full-   1,404     2,803    3,579    3,735    4,108    4,920    
      and part-time    (37%)     (33%)    (33%)    (33%)    (34%)    (33%)    
      employment                                                     
      o  full-time     1,163     2,178    2,852    2,995    3,311    3,927    
                       (30%)     (26%)    (26%)    (26%)    (27%)    (26%)    
      o  part-time     241 (6%)  625 (7%) 727 (7%) 740 (7%) 797 (7%) 993 (7%) 
      o  Total with                                3,594    4,400    5,840    
      any benefits                                 (31%)    (36%)    (38%)    
      o  VA benefits      NA        NA       NA    1,530    2,091    2,924    
      only                                         (13%)    (17%)    (19%)    
      o  other            NA        NA       NA    1,494    1,699    2,089    
      benefits only                                (13%)    (14%)    (14%)    
      o  both VA and      NA        NA       NA    570 (5%) 610 (5%) 827 (5%) 
      other benefits                                                 

Source: GAO analysis of VA data.

Notes: Percentages may not add up to total shown due to rounding. NA =
Data on receipt of VA and other benefits were not available for fiscal
years 2000 through 2002.

Greater Self-Determination in Terms of Improved Functioning in Several
Areas

To track greater self-determination, VA examines such goals as veterans'
progress in handling of alcohol, drug, mental health, and medical problems
and overcoming deficits in social or vocational skills. 20 A greater
proportion of veterans leaving the program each year have met these goals,
with 57 to 69 percent showing improved functioning in fiscal year 2005, as
shown in figure 8.

Figure 8: Percentage of Veterans Leaving the GPD Program with Greater
Self-Determination, Fiscal Years 2000 through 2005

Note: The percentage calculations are based on the number of veterans who
showed the problem at admission.

These improvements have occurred while the proportion of veterans who
entered the GPD program with a history of such problems remained constant
or increased. Specifically, the proportion entering with substance abuse
problems who left the program in fiscal years 2000 through 2005 remained
relatively constant, while the proportion of veterans with a history of
mental or medical illness more than doubled, according to VA data. See
table 4.

Table 4: Number of Veterans Leaving GPD Program and Percentage with
Specific Problems at Entry, Fiscal Years 2000 and 2005

Number of discharges from GPD program    Fiscal year 2000 Fiscal year 2005 
                                            4,020            15,403           
Number of discharges for whom data are   3,826            15,048           
available                                                 
Problems that discharged veterans showed                  
on entering the program:                                  
      o  Alcohol                                 2,789 (73%)     11,180 (74%) 
      o  Drugs                                   2,579 (67%)     10,307 (68%) 
      o  Mental illness                          1,205 (32%)      9,736 (65%) 
      o  Medical illness                         1,255 (33%)     10,488 (70%) 
      o  Social or vocational                    2,276 (60%)     10,864 (72%) 

Source: GAO analysis of data from NEPEC annual reports.

In 2006 VA Took Steps to Help Ensure That VA Liaisons Conduct Required
Reviews of GPD Provider Performance

In addition to assessing the program through the success of its veterans,
VA policy calls for all VA liaisons to review the performance of
individual GPD providers in meeting objectives that are identified in
their grant documents. Providers are required to establish specific
measurable objectives for each of the three program goals. To reach the
housing goal, for example, some providers we visited established savings
objectives, requiring veterans to set aside a portion of any income they
receive so that they can accumulate sufficient cash reserves to cover
costs of renting a room or apartment when they leave the program. Most
providers we visited also set outcome objectives for the percentage of
veterans expected to obtain independent housing when they left the
program. For the income goal, some providers set objectives requiring that
a certain percentage of veterans be offered or enrolled in vocational
training, develop resumes, interview for jobs, or apply for entitlement
benefits. Most providers also set objectives that a certain percentage of
veterans would find work. For the self-determination goal, some providers
required that a certain percentage of veterans maintain sobriety or attend
weekly Alcoholics or Narcotics Anonymous meetings.

In its 2006 examination of the GPD program, VA's OIG found, however, that
many providers had not tracked their performance in achieving these
objectives and some VA liaisons had not reviewed the providers'
performance. The OIG recommended that VA liaisons ensure that the
providers' performance be monitored. The GPD program office has since
moved to enforce the requirement that VA liaisons review GPD providers'
performance when the VA team comes on-site each year to inspect the GPD
facility. 21 The VA liaison will have the flexibility to determine the
method for reviewing and recording the providers' performance, so long as
the results are documented. GPD providers who do not meet performance
objectives will be required to work with their local VA staff to create a
corrective action plan or resubmit their applications with new objectives.

VA Does Not Routinely Collect Data on Veterans' Long-Term Success, but
Recent Study May Provide Insights on How Veterans Fare a Year after
Leaving the Program

VA does not require that veterans be contacted for purposes of program
evaluation after they leave the GPD program. With a view to the long-term
health of veterans, however, VA attempts to have its clinicians provide
GPD participants with a substance abuse or mental health assessment within
2 months of leaving the program. In addition, the forms completed when
veterans leave the GPD program identify any follow-up that may have been
arranged to help them continue to cope with problems that they have
experienced. While follow-up is not required, about 80 percent of GPD
providers reported that they conduct some sort of follow-up with veterans
after they leave the GPD program. Providers may call veterans who have
left, obtain data on those who return for additional support services, or
arrange reunions or other gatherings. Some grant documents also indicate
that the providers planned to measure their performance, in part by
following up with veterans from 3 to 12 months after they left the
program. Some providers follow up to meet the requirements of non-VA
funding they receive. Several providers we interviewed had DOL grants
requiring them to report the employment status of veterans 3 and 6 months
after they left the DOL program. These providers were able to report
results for the veterans deemed employable who participated in both the
GPD and DOL programs. However, GPD participants who were deemed
unemployable because of their disabilities may not have been included in
the DOL program. While many providers attempt to follow up with veterans,
several told us that it is sometimes difficult to maintain contact,
especially with veterans lacking telephones or reliable mailing addresses
and with veterans who have moved away from the area.

While VA considers it important for veterans to achieve immediate success
on leaving the GPD program, homeless veterans may experience setbacks
later on that may negatively affect their housing arrangements, employment
and financial benefits, and self-determination. Furthermore, veterans who
were not immediately successful on leaving the program nevertheless may
have benefited from participating and may be able to achieve success at a
later time. To explore the long-term outcomes of program participants, VA
funded a onetime follow-up study in May 2001 to examine the outcomes for a
randomly selected sample of about 1,300 veterans spread across five
geographic locations who were participating in the GPD program and two
other VA-sponsored homeless programs. According to a VA official, the cost
of the study was about $1.5 million. 22 Included in the sample were 520
veterans housed with 19 GPD providers. Proportionately more veterans in
the GPD programs were chronically homeless, while veterans in one of the
other programs had higher levels of serious medical and psychiatric
problems and greater impairments. At the time of selection, the veterans
had various lengths of stays in these programs.

For the study, university and RAND Corporation researchers interviewed
veterans to determine their status at 1, 3, 6, and 12 months after they
left the programs, with the last interviews conducted in October 2005.
About 360 of the former GPD participants responded to the last interviews.
VA officials do not expect to release final results of the study until
2007, but preliminary results show that just over 80 percent of the GPD
participants had housing 12 months after they left the program. Other
outcomes that are expected to be included in the report are the number of
days that the veterans have either been housed or homeless, their income
and employment situation, their use of drugs and alcohol, their physical
and mental health status, and quality of life.

Conclusions

Addressing homelessness is a daunting challenge, given the difficulties
associated with identifying those who need help and the broad spectrum of
services that need to be successfully tailored, coordinated, and delivered
in order to enable individuals and even families to secure permanent
housing and to live more independently. Limited resources-particularly the
availability of affordable permanent housing-make this job even more
difficult. Moreover, the physical and emotional conditions including
substance abuse, and mental illness, prevalent in the homeless veteran
population further increase the difficulty.

VA has taken a number of steps to tackle this challenge by enhancing its
ability to estimate how many veterans need assistance, increasing the
number of GPD beds, instituting measures that help gauge the program's
effectiveness, and through the GPD program, working proactively with local
and federal government agencies and nonprofits to provide the assistance
needed. However, more could be done to optimize VA's investment,
particularly with respect to ensuring policies and criteria are clearly
understood and consistently applied and assessing longer-term outcomes. In
enhancing communications, VA will need to identify effective ways of
sharing information with the more than 100 agency liaisons in addition to
the 300 local GPD program providers-each with a potentially different
means of operating. In assessing longer-term outcomes, VA will need to
weigh the costs, benefits, and feasibility of implementing a variety of
analytical approaches. Clearly, these endeavors will not be easy, but they
are critical to better equipping VA to help homeless veterans.

Recommendations for Executive Action

We recommend that the Secretary of Veterans Affairs take the following two
steps to improve and evaluate the GPD program:

           1. To aid GPD providers in better understanding the GPD policies
           and procedures, we recommend that VA take steps to ensure that its
           policies are understood by the staff and providers who are to
           implement them. For example, VA could make more information, such
           as issues discussed during conference calls, available in writing
           or online, hold an annual conference, or provide training that may
           also include local VA staff.
           2. To better understand the circumstances of veterans after they
           leave the GPD program, we recommend that VA explore feasible and
           cost-effective ways to obtain such information, where possible
           using data from GPD providers and other VA sources. For example,
           VA could review ways to use the data from its own follow-up health
           assessments and from GPD providers who collect follow-up
           information on the circumstances of veterans whom they have
           served.

Agency Comments and Our Evaluation

We provided a draft of this report to VA for review and comment. VA agreed
with our findings and concurred with our recommendations and provided
information on initiatives it has under way or planned that will address
issues raised in our report as well as other challenges the GPD program
faces.

VA concurred that there is an apparent lack of consistency in GPD program
implementation and stressed its commitment to further enhance
communications with VA liaisons and GPD providers, including providers
whose operations are still in the developmental stage. For example, VA
plans to develop a comprehensive GPD implementation plan that will address
several operational issues, including training and certification
requirements. As well, for the first time, the VA's Veterans Health
Administration plans to host a conference or series of regional
conferences for GPD providers and VA liaisons to review program
requirements and expectations. VA estimates these conferences will take
place in spring 2007.

VA also concurred with the need to better understand the circumstances of
veterans after they leave the GPD program and stated that it has plans in
place to address optional approaches for long-term study in this area
after it completes an analysis of its longitudinal outcome studies of VA's
homeless program. In the interim, VA said it would continue to explore
options for using existing data to evaluate program effectiveness.

However, the agency disagreed with the statement in our draft report that
VA officials attribute the decrease in the estimates of homeless veterans
to VA's estimation process and better local data. VA believes that the
recent decrease in the estimates is a direct result of its progress in
treating these veterans through the GPD program.

Several factors may have contributed to the decrease in the estimates of
homeless veterans. We did not intend to imply that the decrease was solely
attributable to changes in VA's estimation process and better local data,
nor did we intend to downplay VA's program successes. We have revised the
language in this report accordingly.

VA's written comments appear in appendix V. VA also provided technical
comments, which have been incorporated into the report as appropriate.

We are sending copies of this report to the Secretary of Veterans Affairs.
We will also make copies available to others on request. In addition, the
report will be available at no charge on GAO's Web site at
http://www.gao.gov . If you or your staff have any questions about this
report, please contact me at (202) 512-7215 or [email protected] . Contact
points for our Offices of Congressional Relations and Public Affairs can
be found on the last page of this report. GAO staff who made major
contributions to this report are listed in appendix VI.

Sincerely yours,

Cristina T. Chaplain Acting Director Education, Workforce, and Income
Security Issues

Appendix I: Scope and Methodology
The objectives of this report were to review (1) Department of Veterans
Affairs (VA) estimates of the total number of homeless veterans and the
number of transitional beds available, (2) the extent of collaboration
involved in the provision of Homeless Providers Grant and Per Diem (GPD)
program and related services, and (3) VA's assessment of GPD program
performance.

In conducting our review, we focused on the GPD providers that serve the
general homeless veteran population rather than those serving veterans
with special needs, although we visited some special needs grantees. We
interviewed officials at VA headquarters, the GPD program office, the
regional Veterans Integrated Service Networks, VA's Northeast Program
Evaluation Center (NEPEC), and organizations knowledgeable about homeless
veterans' issues, including the National Coalition for Homeless Veterans.
To gain an initial understanding of the GPD program in operation, we spoke
with staff and toured GPD facilities in Baltimore, Maryland; Denver,
Colorado; and Washington, D.C. To develop greater in-depth material for
this report, we made more extensive visits to 13 GPD providers that fall
under the responsibility of VA's medical centers in Boston, Massachusetts;
Los Angeles, California; Tampa, Florida; and Tomah and Madison, Wisconsin.
We selected these GPD providers to obtain a range of geographic locations,
size of programs, and proximity to VA medical centers. (See table 5 for a
listing of sites we visited and their characteristics.) During our visits,
we toured GPD facilities, interviewed GPD providers, medical center staff,
community agencies that partner with the GPD providers, and current and
former GPD program participants. Additionally, we interviewed staff but
did not tour facilities of 16 other GPD providers in the areas we visited.
We also met with GPD and other service providers at conferences sponsored
by the Departments of Labor and Health and Human Services.

Table 5: Features of GPD Programs That GAO Visited

                                 Number of    Fiscal    Fiscal Year Location  
                                  GPD beds  Year `05  `05Discharges Typea     
                                              Admits                
Massachusetts-Boston, Fitchburg, Leominster Veterans Integrated Service
Network 1                    
New England Shelter for              30       149            137 urban     
Homeless Veterans,                                               
Post-Detox Program (Boston)                                      
Veteran Hospice Homestead            12        19             21 rural     
(Fitchburg)                                                      
The Armistice Homesteadb             15        NA             NA rural     
(Leominster)                                                     
Florida-Tampa, Melbourne, Cocoa Veterans Integrated Service Network 8
Agency for Community                 60        64             53 urban     
Treatment Services, (Tampa)                                      
Vietnam Veterans of Brevard          19        70             53 urban     
(Melbourne)                                                      
Volunteers of                        80       100            100 urban     
America-Florida (Cocoa)c                                         
Wisconsin-Tomah, Madison, Fort McCoy Veterans Integrated Service Network
12                           
Veterans Assistance                  60       162            167 rural     
Foundation (Tomah)d                                              
Veterans Assistance                   7         9             10 urban     
Foundation, Step Up Program                                      
(Madison)                                                        
Wisconsin Department of              14        23             18 rural     
Veterans Affairs  (Fort                                          
McCoy)                                                           
California-Los Angeles Veterans Integrated Service Network 22
P.A.T.H.c                            10        28             23 urban     
The Salvation Army, The              95       193            200 urban     
Haven d                                                          
Volunteers of America-LAc           102       106             98 urban     
Weingart Center Association         100       107            113 urban     

Source: GAO review of VA data.

aVA does not classify grantees as rural; however, we included this type of
information for site selection purposes.

bThe Armistice Homestead is part of a collaborative grant under
Massachusetts Veterans Inc. The entire grant funds 43 beds, 15 of which
are located at the Armistice. Specific admission and discharge data were
not available for the Armistice program.

cProgram also has funding for a service center.

dProgram is located on VA medical center grounds.

Throughout our review, we worked with the VA's Office of Inspector General
(OIG) to ensure that we complemented but did not duplicate a review it was
conducting on GPD program management. The OIG's review was designed to
determine if records demonstrate that (1) homeless veterans receive
appropriate assessment and treatment, (2) GPD provider performance is
evaluated and actions are taken to improve conditions, (3) GPD providers
achieve their stated goals, (4) VA's guidelines for the inspection of GPD
facilities are followed, (5) GPD operations are properly monitored by VA,
and (6) fiscal controls are adequate. Although the OIG's report was not
available at the time we prepared our report, we were briefed on results
that were relevant to our work and incorporated the information as
appropriate. In addition, we discussed with the OIG's team our selection
of sites to visit and chose sites that were not included in the team's
review.

In reviewing VA estimates of the number of homeless veterans, we reviewed
the literature, read relevant reports, and interviewed VA officials,
particularly those involved in the federally mandated Community
Homelessness Assessment, Local Education and Networking Group for Veterans
(CHALENG). We interviewed experts in the subject area and officials with
the Bureau of the Census and the Department of Housing and Urban
Development (HUD). We used information from our site visits to supplement
our discussion on how local entities conduct counts of homeless
individuals. We did not review the validity of VA's estimates. To identify
GPD program capacity, location, and number of admissions, we analyzed data
from a series of annual reports prepared by NEPEC, updated where
appropriate by information from the GPD program office in May 2006.

To assess the overall extent to which GPD providers collaborated with
other agencies to offer services to homeless veterans, we analyzed NEPEC
survey data. The survey included responses from all GPD providers in 2003,
when NEPEC first conducted the survey, and all programs that became
operational or were funded in subsequent years through November 2005. For
more information on the survey data, see appendix III. We performed basic
reasonableness tests on the survey data and contacted NEPEC for any
clarifications or discrepancies. We determined these data to be
sufficiently reliable for the purposes of this report. To get an
understanding of how collaboration was actually occurring at the local
level, we conducted site visits. During these visits we gathered
information on the types of services GPD providers offer, how providers
partnered with local agencies (including VA) to offer services, and how
these partnerships were working. To review how VA coordinates with other
federal agencies, we attended a meeting of VA's Advisory Committee on
Homeless Veterans, talked with a representative from the Interagency
Council on Homelessness, and contacted other prominent federal partners.

To identify how VA assesses the performance of the GPD program, we
reviewed GPD program goals, interviewed VA officials, including a team
with the OIG, and analyzed data obtained from VA's national program office
and NEPEC. We reviewed the Grant and Per Diem Program Evaluation
Procedures Manual that NEPEC sends to each VA liaison that describes the
responsibilities of liaisons and GPD providers in completing, reviewing,
and submitting intake and discharge forms on individual participants. We
extracted data on outcomes from tables included in NEPEC's series of
annual reports on the program and discussed the reliability of these data
with NEPEC officials. This information is briefly summarized in appendix
IV along with relevant findings from the OIG's review. We did not
independently verify the NEPEC data. We reviewed how VA collects and
analyzes outcome data and found these data to be sufficiently reliable for
our purposes. Additionally, we reviewed grant documents for the sites we
visited to identify the specific objectives they set to meet program goals
and asked VA officials and providers about various aspects of performance
measurement during our site visits. We did not conduct our own review of
outcomes for homeless veterans served by the GPD providers we visited.

At the time we conducted our analysis, VA's follow-up study had not been
released; therefore, our discussion of the study is based on our review of
preliminary results that identified the numbers and characteristics of the
participants, the timetable and roles of the universities and researchers
involved, and the housing outcomes at the end of the year. Conducted from
2001 through 2005, the study followed a total of 1,294 participants, with
approximately 260 participants from each of five medical center areas
serving California, the District of Columbia, Florida, Maryland, Ohio,
Pennsylvania, and West Virginia. Veterans were randomly selected from
lists of active participants that included recent admissions as well as
participants with longer stays in the program. Participants were drawn
from programs operated by 6 domiciliary care providers, 16 contracted
residential treatment providers, and 19 GPD providers. The study had an
overall response rate of 72 percent for all participants in the three
transitional housing programs, with a response rate of 69 percent for the
GPD participants, for the interviews conducted a year after they left the
program. Of the 520 GPD participants studied, 359 were interviewed a year
after leaving the program. Of those interviewed, 60 percent were in their
own independent housing, 23 percent were sharing with friends or family,
and 15 percent were in temporary housing, including shelters or in an
institution other than a jail.

We conducted our work between August 2005 and July 2006 in accordance with
generally accepted government auditing standards.

Appendix II: VA's Programs for Homeless Veterans Other than the GPD
Program

Veterans Health Administration Programs for Homeless Veterans

Health Care for Homeless Veterans (HCHV) including Contracted Residential
Treatment

Under the HCHV umbrella program, VA provides outreach, health and mental
health assessments, treatment, and referrals for homeless veterans with
mental health and substance abuse problems. Veterans with limited length
of service or with other than a dishonorable discharge are eligible for
the HCHV program but may not necessarily be eligible for VA health care,
where the criteria are more restrictive. A veteran needing transitional
housing while undergoing treatment may be placed in one of the
approximately 300 contracted residential treatment beds that are funded
from the budgets of individual medical centers. In fiscal year 2005, there
were about 1,700 admissions for an average stay of 2 months at $36 per
day; the recommended maximum stay is 6 months. Where contracted
residential treatment is not available, veterans in need of transitional
housing may be referred to the more widely available GPD program or
domiciliary care. In fiscal year 2005, VA's HCHV program provided
outreach, treatment, and referral services to about 61,000 homeless
veterans, with obligations of about $40 million.

Homeless Domiciliary Residential Rehabilitation and Treatment Program

This transitional housing program is designed for homeless veterans who do
not need hospital or nursing home services while their clinical status is
being stabilized. In this program, veterans receive various services,
including medical and mental health evaluations, treatment, and community
support. Domiciliary programs are generally located on the grounds of VA
medical centers, and unlike the GPD programs, they are usually managed and
staffed by the local VA medical center. In fiscal year 2005 about 5,000
homeless veterans stayed an average of 4 months in this program. About
1,800 beds were available exclusively for homeless veterans, with
obligations of about $58 million. Additional funding was awarded in 2005
to increase the number of beds available to about 2,200 in fiscal year
2007, bringing total obligations up to a projected $73 million.

Homeless Compensated Work Therapy/Transitional Residence

This work therapy program provides veterans with job skills and income.
Through the program veterans produce items for sale or provide services
such as temporary staffing to a company. While participating in this
program, veterans may receive individual or group therapy and follow-up
medical care on an outpatient basis. At some locations, program
participants can stay in one of the about 500 beds available in
transitional, community-based group homes. Veterans participating in this
program are required to use a portion of their income from the work
program to pay for rent, utilities, and food. Obligations for this program
in fiscal year 2005 were about $10 million.

Loan Guarantee for Multifamily Transitional Housing

This transitional housing program provides guaranteed loans to nonprofit
organizations to construct or rehabilitate multifamily transitional
housing for homeless veterans, including single room occupancy units.
Supportive services and counseling, including job counseling, must be
provided with the goal of encouraging self-determination among
participating veterans. Veterans must maintain sobriety, seek and maintain
employment, and pay a fee in order to live in these transitional units.
Not more than 15 loans with an aggregate total of $100 million may be
guaranteed under this program. In fiscal year 2005, the Vietnam Veterans
of San Diego housing project was under construction. Other programs have
been conditionally selected and are expected to be approved in fiscal
years 2006 and 2007. For information on the challenges encountered in
implementing this initiative, see Related GAO Products for GAO's report on
this program.

Housing and Urban Development-VA Supported Housing

This permanent, subsidized housing program provides HUD rental assistance
(Section 8) vouchers for use by homeless veterans with chronic mental
health or substance abuse disorders. Veterans are required to pay a
portion of their income for rent; those without income receive fully
subsidized housing. In general, veterans who do not exceed the maximum
allowable income can remain in the housing permanently, but must agree to
intensive case management services from VA staff and make a long-term
commitment to treatment and rehabilitation. Local housing authorities
control access to the vouchers. Many of the 1,780 vouchers allocated by
HUD remain in use but no new vouchers have been made available. As a
result, in fiscal year 2005, only 142 veterans were admitted to the
program. VA's obligations in support of this program in fiscal year 2005
were about $3 million.

Veterans Benefits Administration Programs for Homeless Veterans

Veterans Benefits Administration Outreach

According to VA, in 20 of its 57 regional offices VA has designated
full-time homeless veterans coordinators who work with HCHV and other VA
staff to conduct joint outreach, provide counseling, and offer other
services to homeless veterans, such as helping them apply for veterans
benefits. In the remaining regions, staff may be assigned collateral
responsibility to work with homeless veterans. One of the goals of this
program is to expedite the processing of benefit claims made by homeless
veterans. According to VA, in fiscal year 2005, VA received approximately
4,400 claims from homeless veterans. Of these claims, 56 percent were for
disability compensation and 44 percent were for pensions. Of the
compensation claims, 26 percent were granted, 33 percent denied, and 41
percent pending an average of about 4 months. Of the pension claims, 62
percent were granted, 18 percent denied, and 21 percent pending an average
of about 3 months.

Acquired Property Sales for Homeless Providers

VA properties that are obtained through foreclosures on VA-insured
mortgages are available for sale at below fair market value to nonprofit
and public agencies that use the properties to shelter or house homeless
veterans. Since the inception of this program, more than 200 properties
have been sold or leased.

Labor-VA Incarcerated Veterans' Transition Program

Under this demonstration program, the Department of Labor (DOL) funds
community agencies to provide training and support services, and VA
contributes its services, to help veterans who are incarcerated and at
risk of homelessness make a successful transition back into the workforce.
According to DOL, services provided include career counseling, employment
training, job-search and job-placement assistance, life-skills
development, and follow-up. Local staff from both VA's Health
Administration and Benefits Administration provide information about
available VA benefits and services. Grantees must report the number of
veterans who are still employed 6 months after job placement, whether they
are in the same or similar jobs, and the reasons why veterans who were
placed are no longer employed. DOL provided $2 million to seven community
agencies in 2006 for this purpose.

Appendix III: Range of Services Offered by GPD Programs Nationwide

We analyzed NEPEC's Facility Survey data to identify the types of services
that programs provide and how they are provided. NEPEC conducted the
survey to capture information on the types of GPD programs funded.
According to NEPEC officials, the survey was used to capture information
such as program location, admissions criteria, services available, and
licensing. Because the survey was not intended to be used as a tool to
review how programs were performing, NEPEC does not conduct rigorous
internal reviews of the data collected. We conducted basic reasonableness
tests and contacted NEPEC for any clarifications or discrepancies. We
found the survey data sufficiently reliable for the purposes of this
report.

The survey was first deployed in 2003 to all agencies that were receiving
funding that year. In subsequent years, NEPEC had newly funded agencies
complete this onetime survey. A total of 281 transitional housing
facilities were included in the survey data we analyzed-148 of the
facilities were surveyed in 2003, 94 in 2004, and 39 in 2005. According to
NEPEC, this represents all operational programs as of November 2005. While
there were about 300 agencies with GPD grants, some of the agencies have
multiple grants for one facility, resulting in one survey being completed
for that facility. The surveys were completed by the VA liaisons in
consultation with GPD provider staff. NEPEC officials were confident they
have achieved a 100 percent response rate. While we did not independently
verify the response rate for the survey, we concluded that it would be at
least 90 percent.

Table 6 shows the percentage of facilities that reportedly provide the
selected services and how the services were provided. Survey respondents
were asked to identify how, if at all, services were provided and were
directed to choose only one method. It may be the case, however, that as
in some locations we visited, services were provided by more than one
method. As can be seen, the majority of GPD programs provided a spectrum
of services for veterans. However, these programs varied in how services
were provided, with some services more likely to be provided through
partnerships and others more likely to be provided in-house directly by
staff. Some of the services that were more likely to be provided through
partnerships include those that require counseling or medical-related
treatment. Services primarily provided directly by GPD providers tended to
be more related to case management type activities.

Table 6: Percentage of GPD Facilities Reporting They Provided Selected
Services by Method

                                    How services were provided by  
                                              programs             
Services (ordered by prevalence Indirectly Indirectly  Directly      Total 
of service being offered)          through  by staffb by staffc percentage 
                                    linkagesa                              of 
                                                                   facilities 
                                                                    providing 
                                                                      service 
Vocational/educational                48.0       11.5      39.8       99.3 
counseling                                                      
Discharge planning                     8.2        2.5      88.6       99.3 
Assistance with obtaining             25.7        5.4      67.5       98.6 
social services (e.g.,                                          
Medicaid, Supplemental Security                                 
Income, Social Security                                         
Disability Insurance)                                           
Case management services              10.0        3.2      85.4       98.6 
Housing assistance                    20.8        4.3      73.1       98.2 
Assistance with spending money,       18.6        8.2      70.0       96.8 
banking or other financial                                      
matters                                                         
Transportation or assistance          20.0        7.1      69.3       96.4 
using public transportation                                     
Relapse prevention groups             48.8        5.4      41.9       96.1 
Comprehensive mental health           70.4        8.6      16.8       95.7 
assessment/diagnosis                                            
Individual therapy                    47.9        5.4      42.5       95.7 
Referral to other transitional        13.9        3.2      78.2       95.4 
services                                                        
Comprehensive substance abuse         47.7        8.6      38.7       95.0 
assessment/diagnosis                                            
Group therapy, not including          38.6        5.7      47.1       91.4 
relapse prevention                                              
Aftercare counseling                  48.2        5.4      37.5       91.1 
AIDS screening and counseling         75.0        7.5       6.8       89.3 
Nutritional counseling                54.6       11.4      23.2       89.3 
Legal advice or counseling            76.1        3.9       5.7       85.7 
Outcome follow-up (post               18.6        6.1      56.3       81.0 
discharge)                                                      
Family counseling                     44.3        7.1      28.6       80.0 
Religious or spiritual                51.1        8.2      17.5       76.8 
counseling                                                      
Domestic                              63.2        4.3       8.2       75.7 
violence-family/partner                                         
violence services                                               
Representative payee servicesd        51.4        3.6       8.6       63.6 
Child care                            17.1        3.2       1.8       22.1 

Source: GAO analysis of NEPEC GPD program facility survey.

Note: Percentages were calculated for facilities that completed the survey
question, either 279 or 280 facilities depending on the question.

aIndirectly through linkages means treatment is provided indirectly
through links with other agencies, including VA.

bIndirectly by staff means treatment is provided indirectly by other staff
of the organization.

cDirectly by staff means treatment is provided directly by staff at this
program.

dRepresentative payees handle an individual's benefits if the individual
is unable to. The benefits must be used to meet the needs of the
beneficiary.

Appendix IV: Participant Outcomes for the Grant and Per Diem Program

Outcomes are reported on a standard Northeast Program Evaluation Center
discharge form that must be filled out by VA staff or by GPD staff with
VA's review and sign-off when the participant leaves the program. The form
also captures information on the length and cost of stay in the GPD,
reasons the participant left the program, and any plans for follow-up
treatment for substance abuse or other problems. NEPEC officials told us
that they do not verify the data submitted to them, but they do perform
tests for completeness and internal consistency. VA's Office of Inspector
General (OIG) found that not all outcomes shown on the discharge forms
were supported by additional information in the sample of case records
that the OIG reviewed. For example, 76 percent of records included
information supporting the veterans' outcomes indicated on the form, but
about 24 percent of records lacked such support.

Outcomes for housing and income are shown as a percentage of all
participants who left the program for any reason. However, outcomes for
self-determination in terms of improved functioning are shown as a
percentage of those veterans who had an identified problem when they
entered the program. The determination that a participant has or has not
improved may be considered somewhat subjective. The problems are described
by participants themselves to VA staff in response to a series of
questions on a standard NEPEC intake form that also includes a section for
the VA clinical staff to record their observations of the substance abuse
or mental health problems that the participants face. The intake form also
captures other characteristics of the participants, such as their
military, financial and living circumstances. VA staff are expected to
complete these forms when they first contact homeless veterans but no
later than the veterans' third day with a GPD provider and to forward the
forms to NEPEC. NEPEC reports that it does not receive intake forms for
about 10 percent of participants in the GPD program each year.

Table 7: Number Served by VA's Health Care for Homeless Veterans and Grant
and Per Diem Program and Veterans' Outcomes, Fiscal years 2000 through
2005

Participants served and  Federal fiscal year (October through September of
outcomes                                    year shown)
                                2000     2001    2002    2003    2004    2005 
Number of                  43,082   57,854  61,123  60,970  63,283  61,261 
                                                                              
      o  veterans treated     34,206   46,862  44,296  42,380  42,485  41,111 
      by VA's Health Care                                                     
      for Homeless             4,841   10,137  11,913  12,396  13,509  16,597 
      Veterans' (HCHV)                                                        
      staff                    4,020    8,706  11,098  11,467  12,454  15,403 
      o  intake assessments                                           
      of homeless veterans                                            
      by HCHV staffa                                                  
      o  admissions of                                                
      veterans to GPDs                                                
      o  discharges from                                              
      GPDs                                                            
Days a veteran stays at        91       85      93     110     126     127 
a GPD, on average                                                  
Housing stability           1,163    2,187   3,073   4,590   6,597   8,186 
outcomes:                                                                  
                                 991    2,162   2,731   2,882   3,127   4,003 
Number of discharges                                               
from GPDs with                                                     
                                                                      
      o  independent                                                  
      housing                                                         
      o  placement in                                                 
      halfway house or                                                
      institution such as                                             
      hospital. nursing                                               
      home, or domiciliary                                            
Increased income or         1,404    2,803   3,579   3,735   4,108   4,920 
skills outcomes:                                                           
                                  NA       NA      NA   2,100   2,701   3,751 
Number of discharges                                                       
from GPDs with                 NA       NA      NA   2,064   2,309   2,916 
                                                                      
      o  full-time or                                                 
      part-time employment                                            
      o  VA benefitsb                                                 
      o  Other public                                                 
      benefitsb                                                       
Greater                     38-42    42-49   43-50   56-62   60-67   62-69 
self-determination                                                         
outcomes:                   43-46    40-44   43-46   50-57   55-63   57-64 
                                                                              
Percentage of discharges       30       32      38      43      49      50 
from GPDs with                                                     
                                                                      
      o  improved alcohol,                                            
      drug, mental healthc                                            
      o  improved medical,                                            
      social/vocational                                               
      conditionc                                                      
      o  success in program                                           

Source: VA data.

aIntake assessments are completed by HCHV staff when they first encounter
a homeless veteran, unless the contact is casual and no services are
offered or referrals made. After a year, new assessments are required if
VA care or services are provided and VA staff have not been working with
the veteran.

bNumbers shown here include veterans who receive both types of benefits as
well as those who receive only the designated benefits. For this reason,
they differ from the numbers shown in table 3.

cPercentages are ranges showing the highest and lowest of each of two or
three outcome measures.

Appendix V: Comments from the Department of Veterans Affairs

Appendix VI: GAO Contact and Staff Acknowledgments

GAO Contact

Cristina T. Chaplain, Acting Director, (202) 512-7215, [email protected]

Acknowledgments

Shelia Drake, Assistant Director; Patricia L. Elston; David Forgosh; and
Nyree M. Ryder made significant contributions to this report. In addition,
Roger Thomas provided legal assistance; Walter Vance and Lynn Milan
analyzed and assessed the reliability of data; Lily Chin, Jonathan
McMurray, and Charles Willson assisted in report development; and Amy
Sheller supported the team during its Los Angeles site visit.

Related GAO Products

Homeless Veterans: Job Retention Goal Under Development for DOL's Homeless
Veterans' Reintegration Program. GAO-05-654T . Washington, D.C.: May 4,
2005.

Veterans Affairs Homeless Programs: Implementation of the Transitional
Housing Loan Guarantee Program. GAO-05-311R . Washington, D.C.: March 16,
2005.

VA Health Care: VA Should Expedite the Implementation of Recommendations
Needed to Improve Post-Traumatic Stress Disorder Services. GAO-05-287 .
Washington, D.C.: February 14, 2005.

Decennial Census: Methods for Collecting and Reporting Data on the
Homeless and Others without Conventional Housing Need Refinement.
GAO-03-227 . Washington, D.C.: January 17, 2003.

Homelessness: Improving Program Coordination and Client Access to
Programs. GAO-02-485T. Washington, D.C.: March 6, 2002.

Homeless Veterans: VA Expands Partnerships, but Effectiveness of Homeless
Programs Is Unclear. GAO/ T-HEHS-99-150 . Washington, D.C.: June 24, 1999.

Homeless Veterans: VA Expands Partnerships, but Homeless Program
Effectiveness Is Unclear. GAO/ HEHS-99-53 . Washington, D.C.: April 1,
1999.

Homelessness: Overview of Current Issues and GAO Studies. GAO/
T-RCED-99-125 . Washington, D.C.: March 23, 1999.

Homelessness: Demand for Services to Homeless Veterans Exceeds VA Program
Capacity. GAO/ HEHS-94-98 . Washington, D.C.: February 23, 1994.

(130513)

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 ( 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
www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

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

Congressional Relations

Gloria Jarmon, Managing Director, [email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Washington,
D.C. 20548

Public Affairs

Paul Anderson, Managing Director, [email protected] (202) 512-4800 U.S.
Government Accountability Office, 441 G Street NW, Room 7149 Washington,
D.C. 20548

www.gao.gov/cgi-bin/getrpt? GAO-06-859 .

To view the full product, including the scope

and methodology, click on the link above.

For more information, contact Cristina T. Chaplain at (202) 512-7215 or
[email protected].

Highlights of GAO-06-859 , a report to the Chairman, Committee on
Veterans' Affairs, House of Representatives

September 2006

HOMELESS VETERANS PROGRAMS

Improved Communications and Follow-up Could Further Enhance the Grant and
Per Diem Program

VA estimates that on a given night about 194,000 veterans were homeless in
2005. The estimate, generally lower than the numbers reported prior to
2004, is considered by VA officials to be the best available. VA officials
believe that its new estimation process and use of better local data have
improved the estimate. While VA has increased the capacity of the GPD
program over the past several years, VA reports that an additional 9,600
transitional housing beds from various sources are needed to meet current
demand. VA has plans to make 2,200 additional GPD beds available.

Number of GPD Beds and Admissions from Fiscal Year 2000 through 2005

GPD providers collaborate with other agencies to help veterans regain
their health and obtain housing, jobs, and various services to enable them
to live independently. However, resource and communications gaps may stand
in the way of VA and provider efforts to meet these goals. Limited
availability of affordable permanent housing, for example, may make it
difficult to move veterans out of homelessness, according to GPD
providers. We also identified instances of misunderstandings of program
policies related to eligibility and program stay limits that could prevent
homeless veterans from being admitted into the GPD program.

VA assesses overall program performance by the success of veterans in
attaining stable housing, income, and self-determination at the time they
leave the program. VA data show that the percentage of veterans achieving
these goals has generally increased or held steady over time. In 2006, VA
also stepped up its assessment of the performance of GPD providers. While
these assessments do not indicate how veterans fare after they leave the
program, preliminary results of a onetime VA study indicate positive
housing outcomes were maintained 1 year later. However, VA does not
routinely collect follow-up data and may not be able to determine how
veterans who were not included in the study are faring after they leave
the program.

About one-third of the nation's adult homeless population are veterans,
according to the Department of Veterans Affairs (VA). Many of these
veterans have experienced substance abuse, mental illness, or both. The
VA's Homeless Providers Grant and Per Diem (GPD) program, which is up for
reauthorization, provides transitional housing to help veterans prepare
for permanent housing. As requested, GAO reviewed (1) VA homeless veterans
estimates and the number of transitional housing beds, (2) the extent of
collaboration involved in the provision of GPD and related services, and
(3) VA's assessment of GPD program performance.

GAO analyzed VA data and methods used for the homeless estimates and
performance assessment, and visited selected GPD providers in four states
to observe the extent of collaboration.

What GAO Recommends

To further strengthen VA's ability to help homeless veterans, GAO is
recommending that VA take steps to ensure policies are understood by
providers and staff who implement them. GAO also recommends that VA
explore feasible and cost-effective means of obtaining information on
long-term outcomes for veterans who leave the GPD programs. VA generally
agreed with our findings and recommendations.
*** End of document. ***