HUD and Treasury Programs: More Information on Leverage Measures'
Accuracy and Linkage to Program Goals Is Needed in Assessing
Performance (18-JAN-08, GAO-08-136).
This is the second of two reports on the leveraging of federal
funds in housing and community and economic development programs.
Leveraging involves using a source of funds to attract other
funds or combining multiple sources of funds. This report
examines (1) the leverage measures and the transparency of the
data and methods used to calculate them, and (2) the relevance of
such measures in assessing performance that the Department of
Housing and Urban Development (HUD) and the Department of the
Treasury (Treasury) reported for six selected programs. To
complete this work, GAO reviewed agency policies and reports,
interviewed officials, and analyzed agency data.
-------------------------Indexing Terms-------------------------
REPORTNUM: GAO-08-136
ACCNO: A79974
TITLE: HUD and Treasury Programs: More Information on Leverage
Measures' Accuracy and Linkage to Program Goals Is Needed in
Assessing Performance
DATE: 01/18/2008
SUBJECT: Data integrity
Economic development
Federal funds
Financial management
Funds management
Government information dissemination
Housing
Housing programs
Information disclosure
Interagency relations
Internal controls
Performance appraisal
Performance measures
Program evaluation
Program management
Reporting requirements
Leverage (financial)
Program goals or objectives
Transparency
Community Development Financial
Institutions Fund
HUD Community Development Block Grant
Program
HUD Home Investment Partnership Program
HUD HOPE VI Program
New Markets Tax Credit Program
Treasury Low Income Housing Tax Credit
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-08-136
Contents:
Letter:
Results in Brief:
Background:
Reported Leverage Measures Lacked Transparency Because Agencies
Generally Did Not Disclose Data Limitations or Calculation Methods:
Leverage Measures Provide Basic Financial Information, but the Extent
to Which They Are Relevant for Assessing Program Performance Varies:
Conclusions:
Recommendations for Executive Action:
Agency Comments and Our Evaluation:
Appendix I: Objectives, Scope, and Methodology:
Appendix II: Transactions Structures for the Selected Treasury
Programs:
Appendix III: Agency-reported Leverage Measures and Our Recalculations:
Appendix IV: Housing and Community and Economic Development Project
Profiles:
Appendix V: Comments from the Department of Housing and Urban
Development, Office of Community Planning and Development:
Appendix VI: Comments from the Department of the Treasury, Community
Development Financial Institutions Fund:
Appendix VII: GAO Contact and Staff Acknowledgments:
Related GAO Products:
Tables:
Table 1: Limitations of Agencies' Leveraging Data and Potential Effect
on Reported Leverage Measures:
Table 2: HUD's Use of Leverage Measures in Performance Assessment and
Other Documents:
Table 3: Treasury's Use of Leverage Measures in Performance Assessment
and Other Documents:
Table 4: Agency-reported Leverage Measures by Program:
Figures:
Figure 1: Calculation Scenarios for a Hypothetical Program:
Figure 2: Performance Measurement Model:
Figure 3: CDFI Transaction Structure:
Figure 4: Low-Income Housing Tax Credit Syndication Transaction
Structure:
Figure 5: New Markets Tax Credit Basic Transaction Structure:
Figure 6: New Markets Tax Credit Leveraged Transaction Structure:
Figure 7: Recalculated Leverage Measures for the CDBG Program:
Figure 8: Recalculated Leverage Measures for the HOME Program:
Figure 9: Recalculated Leverage Measures for the HOPE VI Program:
Figure 10: Institutional and Project Leverage in the CDFI Program:
Figure 11: Institutional and Project Leverage in the New Markets Tax
Credit Program:
Figure 12: Hilltop Oaks Apartments, San Antonio, Texas:
Figure 13: Near North Apartments, Chicago, Illinois:
Figure 14: New Columbia, Portland, Oregon:
Figure 15: ACCION Texas, San Antonio, Texas:
Figure 16: Avenue North, Philadelphia, Pennsylvania:
Figure 17: Gerding Theater at the Armory, Portland, Oregon:
Figure 18: New Hampshire Community Loan Fund Manufactured Housing
Program: Exeter River Cooperative, Exeter, New Hampshire:
Figure 19: Seattle Cooperative Children's Center, Seattle, Washington:
Figure 20: Shoalwater Bay Wellness Center, Tokeland, Washington:
Abbreviation:
CDBG: Community Development Block Grant:
CDE: Community Development Entity:
CDFI: Community Development Financial Institution:
GPRA: Government Performance and Results Act of 1993:
HOME: HOME Investment Partnerships:
HUD: Department of Housing and Urban Development:
IDIS: Integrated Disbursement and Information System:
LEED: Leadership in Energy and Environmental Design:
OMB: Office of Management and Budget:
PART: Program Assessment Rating Tool:
PHA: public housing agency:
SAAHC: San Antonio Alternative Housing Corporation:
Treasury: Department of the Treasury:
United States Government Accountability Office:
Washington, DC 20548:
January 18, 2008:
The Honorable Maxine Waters:
Chairwoman Subcommittee on Housing and Community Opportunity:
Committee on Financial Services:
House of Representatives:
Dear Madam Chairwoman:
In a period of increasingly tight budgets for federal housing and
community and economic development programs, congressional, executive,
and agency decision makers have focused on how best to distribute
scarce federal resources to achieve the greatest benefits--in
particular, the extent to which federal programs leverage private and
other public funds. In response, federal agencies often cite leverage
measures in strategic planning, performance, and budget reports, and on
their Web sites to demonstrate how successful they have been at
attracting other funds to carry out program goals.[Footnote 1]
Under the provisions of the Government Performance and Results Act of
1993 (GPRA), federal agencies are required to measure and report the
performance of their programs.[Footnote 2] GPRA was designed to inform
congressional and executive decision making by providing objective
information on the relative efficiency and effectiveness of federal
programs and spending. A key provision of the act is to create closer
and clearer links between the process of allocating scarce resources
and the expected results to be achieved with these resources, which can
increase the government's capacity to assess competing claims for
federal dollars. Under GPRA, agencies also must complete strategic
plans in which they define their missions, establish results-oriented
goals, and identify strategies to achieve those goals; prepare annual
performance plans that articulate goals aligned with long-term
strategies; and issue annual performance reports in which they report
on actions taken to achieve these goals.[Footnote 3] However, federal
agencies have faced challenges in identifying program goals and
performance measures that go beyond summarizing program activities--for
example, the number of clients served--to distinguishing desired
outcomes or results--for example, improving economic self-sufficiency
among clients served.[Footnote 4]
Further, the current administration has made integrating performance
information into budget deliberations a priority under the President's
Management Agenda.[Footnote 5] The Program Assessment Rating Tool
(PART), which the Office of Management and Budget (OMB) designed, is a
central element of this initiative and consists of a standard series of
questions meant to serve as a diagnostic tool. PART draws on available
program performance and evaluation information, including leverage
measures, to form conclusions about program results and develop follow-
on actions intended to improve those results. As we have reported
previously, access to credible information on program performance is
critical to the success of any program assessment effort, including
PART.[Footnote 6]
As discussed in a May 2007 report on leveraging federal funds for
housing and community and economic development, leveraging can be
defined in two ways: (1) using a relatively small amount of federal
funds to attract private investment and (2) combining or layering
program funds with other federal, state, local, and private sources of
funds.[Footnote 7] Leveraging also can occur at the institutional or
project level--at the institutional level, an entity pools funds from
multiple sources, which later are used to finance a portfolio of
projects; at the project level, an entity leverages funds as necessary
for discrete projects. Further, while leveraging may be useful and
stretch scarce resources, the extent of its use can depend upon local
economic conditions and may have unintended consequences, such as the
substitution of federal funds for private funds that otherwise would
have been contributed to a program or project. Despite differences in
how and under what circumstances programs leverage, little scrutiny has
been placed on the leverage measures these programs report and how
agencies, OMB, and others use such measures to assess performance.
This is the second of two reports undertaken in response to your
request that we examine leveraging as it relates to federal housing and
community and economic development programs.[Footnote 8] For this
report, we examined the Department of Housing and Urban Development's
(HUD) Community Development Block Grant (CDBG), HOME Investment
Partnerships (HOME), and HOPE VI programs and the Department of the
Treasury's (Treasury) Community Development Financial Institutions
(CDFI) Financial Assistance, Low-Income Housing Tax Credit, and New
Markets Tax Credit programs.[Footnote 9] Specifically, this report
examines (1) the leverage measures HUD and Treasury reported for the
selected housing and community and economic development programs and
the transparency of the data and methods used to calculate them and (2)
the relevance of leverage measures in assessing the performance of the
selected programs. This report also provides examples of how federal
funds have been leveraged in the selected programs (see app. II).
To examine the leverage measures HUD and Treasury reported for each of
the selected programs and the transparency of the data and methods used
to calculate them, we reviewed relevant program regulations and
guidance, our prior reports and reports of others, and interviewed
agency officials and other stakeholders. Based on this information, we
requested from HUD and Treasury data they use to measure the extent of
leveraging (for example, data on sources and amounts of funds, or other
financial data, commonly referred to as "leveraging data") in the CDBG,
HOME, and HOPE VI programs and the CDFI and New Markets Tax Credit
programs, respectively, and assessed their reliability in accordance
with our standards.[Footnote 10] Because the Low-Income Housing Tax
Credit program does not have a single, complete source of data on the
extent of leveraging, we surveyed the housing finance agencies--those
organizations that are responsible for administering the program--to
determine what data they collect on the extent of leveraging that
occurs in the program. To examine the relevance of leverage measures in
assessing the performance of the selected housing and community and
economic development programs, we reviewed our and OMB's reports on
performance measurement; agency strategic plans and annual performance
plans, performance and accountability reports, and budget
justifications; and industry and other literature such as agency
reports, press releases, and Web sites.[Footnote 11] We also
interviewed federal agency officials and other individuals with
knowledge of or experience in housing and community and economic
development. As part of this work, we also conducted site visits in
five states and collected information on how federal funds have been
leveraged for a number of projects or initiatives that received funding
from the programs included in our review. Appendix I contains a more
detailed description of our scope and methodology. We conducted this
performance audit in Chicago, Illinois; San Antonio and Laredo, Texas;
Philadelphia and Chester, Pennsylvania; Portland and Salem, Oregon;
Seattle and Tokeland, Washington; and Washington, D.C., from November
2006 to January 2008 in accordance with generally accepted government
auditing standards. Those standards require that we plan and perform
the audit to obtain sufficient, appropriate evidence to provide a
reasonable basis for our findings and conclusions based on our audit
objectives. We believe that the evidence obtained provides a reasonable
basis for our findings and conclusions based on our audit objectives.
Results in Brief:
The leverage measures HUD and Treasury reported for the selected
programs lacked transparency because the agencies generally did not
disclose the limitations of the data or the methods used to calculate
them. We found that for reasons including incomplete reporting of data,
the measures HUD and Treasury reported for the CDBG, HOME, HOPE VI,
CDFI, and New Markets Tax Credit programs did not reflect the actual
extent of leveraging that occurred in the programs. For example,
Treasury lacked leveraging data for approximately 26 percent of New
Markets Tax Credit projects, which could potentially result in an
underestimation of the leveraging that occurred in the program.
However, we found that when the agencies reported leverage measures for
the selected programs in performance and budget reports, and in other
sources, they neither fully disclosed these data limitations, nor
consistently disclosed the method they used to calculate the measures.
Based on our discussions with agency officials, we found that the
agencies generally reported leverage measures that described the ratio
of all other funds (other federal, state, local, and private funds) to
program funds. However, these measures can be calculated in multiple
ways to present different results, such as the extent to which federal
funds are used with nonfederal funds. There is no agency-specific or
governmentwide guidance on what agencies should disclose about the
leverage measures they report or how they calculate them for the
selected programs. We previously have reported that clearly
communicating data limitations and their potential impact may foster
appropriate use of data. Absent specific information on how the
agencies calculated reported leverage measures for the selected
programs and the limitations of those measures, decision makers do not
have sufficient information to understand their meaning and how they
can and should be used in performance assessment, budgeting, and other
contexts.
Leverage measures can provide basic financial information about the
programs included in our review; however, their relevance in assessing
the performance of the selected programs varies considerably. For all
programs, leverage measures can describe inputs, or the resources used
to support program activities, and may be useful for conveying basic
financial information. To the extent that leveraging is a goal or core
(expected) activity of a program, leverage measures generally can
describe program outputs, or the products or services delivered (such
as total leveraged funds) and may be used with other performance
indicators to assess the efficiency and effectiveness of a program in
meeting its goals. Among the programs we reviewed, leveraging is
directly linked to the goals and activities of the CDFI, Low-Income
Housing Tax Credit, and New Markets Tax Credit programs. Each of the
three programs was designed to leverage in multiple ways--the CDFI
program requires CDFIs to leverage additional funds as a condition of
receiving program funds, while the tax credit programs automatically
generate private investment for housing and community and economic
development activities. As a result, OMB and Treasury's use of leverage
measures to describe and assess the performance of these programs
generally was appropriate. In contrast, leveraging is not linked
directly to the program goals and core activities of the selected HUD
programs (CDBG, HOME, and HOPE VI). Leveraging may be a strategy some
funding recipients employ, either by choice or out of necessity, to
meet these programs' goals. Thus, using leveraging to assess impact or
success in meeting goals may create adverse or conflicting incentives
for the agency and funding recipients; for example, giving funding
priority to projects that leverage more over those that leverage less,
but which may fill a greater or more immediate need within a community.
Specifically, emphasizing the importance of leveraging in a program
that provides housing for low-income communities could result in
providing relatively more federal funding to projects that serve higher-
income households and less funding to needier communities, which may
experience difficulty in attracting other funding. Despite the limited
relevance of leveraging to the goals of the CDBG, HOME, and HOPE VI
programs, we found that OMB and HUD often cited leverage measures for
the programs in performance-and budget-related reviews and documents,
including PART reviews.
To ensure that leverage measures provide accurate, relevant, and useful
information to Congress and others, this report makes recommendations
to the Secretaries of HUD and Treasury to disclose information on the
completeness and accuracy of the data and the methods used to calculate
such measures, and if used as a performance indicator, how such
measures link to program goals and core activities. This report further
recommends that the Director of OMB provide guidance to help agencies
determine how to calculate, describe, and use leverage measures in a
manner that is consistent with their programs' design, and re-evaluate
the use of such measures and disclose their relevance to program goals
and activities in future PART or other performance reviews of the
selected programs.
We received written comments on a draft of this report from HUD and
Treasury, which are included in appendixes V and VI, respectively. We
also provided a draft of this report to OMB for review, but no comments
were provided. In a letter from the Acting Deputy Assistant Secretary
for Grant Programs, HUD noted that it was pleased with the results
presented in our draft report, but provided several detailed comments
on and suggested changes to our findings related to the CDBG and HOME
programs (see app. V). For example, HUD expressed concern that the
draft report did not sufficiently emphasize that the CDBG and HOME
programs do not have statutory or regulatory leveraging requirements or
that the agency currently does not publish a leverage measure for the
CDBG program. We incorporated language into the report to address these
comments. In addition, HUD said that it would work to improve the
quality of leveraging data CDBG grantees report to the agency, an
effort that would, in part, address one of our recommendations to the
agency. HUD's Office of Public and Indian Housing also provided
technical comments related to the HOPE VI program, which we
incorporated as appropriate.
In a letter from the Director of the Community Development Financial
Institutions Fund, Treasury expressed appreciation for our finding that
each of the agency's programs included in our review was designed to
leverage. Although Treasury did not specifically comment on our
recommendations, it provided several detailed comments related to the
agency's calculation of leverage measures for the CDFI and New Markets
Tax Credit programs (see app. VI). For example, while Treasury agreed
with our description of the limitations of the data used to calculate
leverage measures for the CDFI and New Markets Tax Credit programs, it
stated that the calculated measures provided reasonable approximations
of the leveraging that occurs in the programs despite these
limitations. However, we continue to believe that these limitations
(which are described in our report and Treasury's comment letter)
potentially could have an impact on the accuracy of the leverage
measures Treasury calculated for the programs and thus should be
adequately disclosed. Accordingly, we did not change the report in
response to these comments. Consistent with our findings and
recommendations, in its comment letter, Treasury acknowledged the
importance of disclosing data limitations and calculation
methodologies, noting that it has done so on numerous occasions with
respect to the CDFI program and stating that it would make every effort
to include such information in any publication of a leverage measure
for the New Markets Tax Credit program.[Footnote 12] HUD's and
Treasury's comments are discussed in greater detail at the end of this
letter.
Background:
HUD's CDBG, HOME, and HOPE VI programs and Treasury's CDFI, Low-Income
Housing Tax Credit, and New Markets Tax Credit programs are among a
number of federal programs that fund housing and community and economic
development. In varying degrees, these programs leverage other funds to
help finance their initiatives and projects. As we reported in a May
2007 report, some of these programs define leveraging as using one
source of funds to attract additional sources of funds, while others
define leveraging more broadly as the layering or combining of
different sources of funds.[Footnote 13] Further, and as described
below, some of these programs leverage at the institutional and project
levels, while some leverage only at the project level. At the
institutional level, an organization (such as a group of investors or a
community or other development authority) pools funds from multiple
sources, which are then used to finance a portfolio of projects. At the
project level, an organization (such as a state or local agency)
leverages funds as necessary to finance discrete projects.
Housing and Community and Economic Development Program Overviews:
We highlight below the purpose, structure, and activities of the three
HUD programs and three Treasury programs that we reviewed. Appendix II
describes in more detail how leveraging occurs in the selected Treasury
programs.
CDBG Program:
The CDBG program is the federal government's principal community
development program. It provides annual grants on a formula basis to
entitlement communities--principal cities of metropolitan statistical
areas, other metropolitan cities with populations of at least 50,000,
and qualified urban counties--and states to develop viable urban
communities by providing decent housing and a suitable living
environment, and by expanding economic opportunities, principally for
low-and moderate-income persons.[Footnote 14] Under the CDBG program,
communities and states develop their own programs and funding
priorities. However, all funded activities must meet one of three
national objectives: primarily benefit low-and moderate-income persons,
aid in the prevention or elimination of slums and blight, or meet
community development needs of particular urgency (because existing
conditions pose a serious and immediate threat to the health or welfare
of the community and other financial resources are not available to
meet such needs). Although the CDBG program has no statutory or
regulatory leveraging requirement, some projects funded under the
program use CDBG funds to leverage additional funds to finance
development costs. In fiscal year 2007, Congress appropriated
approximately $3.7 billion to the CDBG program for formula
distribution, and HUD allocated these funds to 1,133 entitlement
communities, 49 states, and Puerto Rico.[Footnote 15]
HOME Program:
HOME provides formula grants to states and localities--certain cities,
counties, or consortiums of cities and counties--to fund a wide range
of activities to benefit low-income people.[Footnote 16] Under the HOME
program, states and localities may use program funds to finance a broad
range of activities, such as providing eligible homeowners and new
homebuyers with home purchase or rehabilitation financing assistance
and building or rehabilitating housing for rent or ownership. States
and localities also may use HOME funds to provide tenant-based rental
assistance.[Footnote 17] The program requires states and localities to
match 25 percent of expended program funds with monetary or certain in-
kind contributions, such as donated materials or voluntary
labor.[Footnote 18] This match requirement was designed to elicit local
resources in support of affordable housing. Like the CDBG program, the
HOME program has no statutory or regulatory leveraging requirement;
however, some projects funded under the program use HOME funds to
leverage additional funds to finance development costs. In fiscal year
2007, Congress appropriated approximately $1.8 billion to the HOME
program, and HUD allocated these funds to 589 localities, the 50
states, and Puerto Rico.
HOPE VI Program:
HOPE VI is part of HUD's effort to transform public housing.[Footnote
19] By providing funds for a combination of capital improvements and
community and supportive services, the HOPE VI revitalization grant
program seeks to (1) improve the living environment for residents of
severely distressed public housing through the demolition,
rehabilitation, reconfiguration, or replacement of obsolete units; (2)
revitalize sites on which such severely distressed public housing is
located, and contribute to the improvement of the surrounding
neighborhood; (3) provide housing that will avoid or decrease the
concentration of very-low income families; and (4) build sustainable
communities. Any public housing agency (PHA) that has severely
distressed public housing units in its inventory is eligible to apply
for a HOPE VI revitalization grant. Recipients of revitalization grants
must match 5 percent of the grant with other funds, and HUD awards PHAs
that demonstrate an ability to leverage additional points in the HOPE
VI application process.[Footnote 20] In fiscal year 2006, HUD made four
HOPE VI revitalization grants totaling approximately $72
million.[Footnote 21]
CDFI Program:
Through the CDFI program, Treasury's CDFI Fund provides CDFIs with
financial assistance in the form of grants, loans, equity investments,
and deposits to enhance their ability to make loans and investments and
provide services for the benefit of designated investment areas,
targeted populations, or both.[Footnote 22] CDFIs must match (leverage)
their financial assistance awards dollar-for-dollar with funds of the
same type (equity investment, loan, deposit, or grant) from nonfederal
sources.[Footnote 23] CDFI funds can be used for economic development
(job creation, business development, and commercial real estate
development), affordable housing (housing development and
homeownership), and community development financial services (provision
of basic banking services to underserved communities and financial
literacy training). In 2007, Treasury made approximately $26 million in
financial assistance awards to 49 CDFIs.[Footnote 24]
Low-Income Housing Tax Credit Program:
Under the Low-Income Housing Tax Credit program, states are authorized
to allocate federal tax credits to private investors as an incentive to
develop rental housing for low-income households.[Footnote 25] The
equity generated by the sale of the credits is used to lower the
financing costs of housing developments by reducing the debt or equity
the developer otherwise would incur or contribute.[Footnote 26]
Investors who purchase the tax credits may claim the credits annually
for 10 years. To receive Low-Income Housing Tax Credit financing,
properties must meet certain rent and tenant income requirements: (1)
at least 20 percent of the units in the property must be reserved for
individuals or families with incomes of 50 percent or less of the area
median income, or at least 40 percent of the units must be reserved for
individuals or families with incomes of 60 percent or less of the area
median income; and (2) rents for affordable units are restricted to 30
percent of the applicable income limit (that is, 50 percent or 60
percent of the area median income). Each state receives an allocation
of the greater of $1.75 per capita or $2 million annually, adjusted by
a cost of living factor ($1.95 or $2.275 million in 2007).[Footnote 27]
The program costs the federal government an estimated $5 billion
annually in forgone tax revenue and is the government's largest housing
production program.
New Markets Tax Credit Program:
The New Markets Tax Credit program permits taxpayers to receive a
credit against federal income taxes for making qualified equity
investments in designated Community Development Entities (CDE), which
must in turn make investments in low-income communities.[Footnote 28]
Qualified low-income community investments include (1) any capital or
equity investment in, or loan to, any qualified, active, low-income
community business; (2) the purchase from another CDE of any loan made
by such entity that is a qualified low-income community investment; (3)
financial counseling and other services to businesses located in, and
residents of, low-income communities; and (4) certain equity
investments in, or loans to, a CDE. The credit provided to the investor
totals 39 percent of the cost of the investment and is claimed over a 7-
year period. In addition, Treasury scores those applications in which
CDEs demonstrate an ability to leverage additional funds more
favorably.[Footnote 29] In fiscal year 2007, Treasury awarded $3.9
billion in New Markets Tax Credits (totaling approximately $1.5 billion
in forgone federal tax revenue) to 61 CDEs.
Reported Leverage Measures Lacked Transparency Because Agencies
Generally Did Not Disclose Data Limitations or Calculation Methods:
The leverage measures (such as ratios) HUD and Treasury reported for
the selected programs in performance, budget, and other documents
lacked transparency because the agencies generally did not disclose the
limitations of the data or the methods used to calculate them. Based on
our review of available leveraging data and interviews with HUD and
Treasury officials, we found that the leverage measures the agencies
reported for the selected programs were based on incomplete data and
thus did not capture the actual extent of leveraging in the programs.
We also found that while the agencies generally reported measures that
described the ratio of all other funds (federal, state, local, and
private funds) to program funds, alternative measures that described
the total federal investment or total private investment in a program
provided considerably different results--also potentially of value to
decision makers--about the extent of leveraging in a program. Further,
no agency-specific or governmentwide guidance directs what agencies
should disclose about the leverage measures they report for the
selected programs; however, we regularly have reported that clearly
communicating data limitations and their potential impact may foster
appropriate use of data.[Footnote 30] Consequently, absent specific
information on how these measures were calculated and their
limitations, decision makers would not have sufficient information to
understand their meaning and determine how they could and should be
used in performance assessment, budgeting, and other contexts.
HUD and Treasury Did Not Always Disclose Limitations of Data Used to
Determine Leverage Measures for the Selected Programs:
Based on our review of available leveraging data and interviews with
HUD and Treasury officials, we found that the leverage measures the
agencies reported for the selected programs were based on incomplete
data and did not capture the actual extent of leveraging that may have
occurred in each of the programs.[Footnote 31] We also found that HUD
and Treasury did not always disclose these limitations when they
published the measures. Table 1 describes the limitations associated
with the underlying data used for determining leverage measures for
each of the selected programs.
Table 1: Limitations of Agencies' Leveraging Data and Potential Effect
on Reported Leverage Measures:
Program(s): CDBG and HOME;
Data limitation: The database HUD used to collect leveraging data for
the programs did not distinguish between nonresponses, which default to
zero, and actual entries of zero (that is, $0).[B];
Effect of data limitation on leverage measure[A]: Assuming that some
grantees failed to enter funding information (which would appear in the
data as $0), the total amount of leveraging that occurred in each
program potentially would be underestimated.
Program(s): HOPE VI;
Data limitation: HUD's database captures data on leveraging that
occurred in completed phases of HOPE VI developments rather than data
on leveraging that occurred in completed HOPE VI developments (which
comprise multiple phases of development);
Effect of data limitation on leverage measure[A]: To the extent that
HOPE VI developments included in the calculation did not include all
phases of construction, the total amount of leveraging that occurred in
the program potentially would be underestimated.
Program(s): CDFI;
Data limitation: In its calculation of institutional leverage, Treasury
assumed match leverage--that is, the ratio of nonfederal match funds to
program funds--to be 1 to 1. According to Treasury officials, CDFIs may
attract more than $1 in nonfederal funds for every $1 received in
program funds; however, the agency does not collect data on match
contributions that exceed the $1 requirement;
Effect of data limitation on leverage measure[A]: To the extent that
match funds exceeded the reported 1 to 1 ratio, the total amount of
leveraging that occurred in the program potentially would be
underestimated.[C].
Program(s): Low-Income Housing Tax Credit;
Data limitation: Treasury does not collect leveraging data or report a
leverage measure for the program.[D];
Effect of data limitation on leverage measure[A]: Not applicable.
Program(s): New Markets Tax Credit;
Data limitation: Treasury assumed that CDEs contribute 100 percent of
available tax credit equity to qualified low-income community
investments even though program regulations permit CDEs to retain up to
15 percent of the equity for administrative and other purposes.[E];
Effect of data limitation on leverage measure[A]: To the extent one or
more CDEs contributed less than 100 percent of available tax credit
equity to qualified low-income community investments, the total amount
of leveraging that occurred in the program potentially would be
overestimated.
Program(s): New Markets Tax Credit;
Data limitation: Program(s)Data limitation: Project-level data were
unavailable for 26 percent of the projects funded under the program;
Effect of data limitation on leverage measure[A]: Program(s)Effect of
data limitation on leverage measure[A]: To the extent projects for
which data were unavailable leveraged additional funds at the project
level, the total amount of leveraging that occurred in the program
potentially would be underestimated.
Source: GAO analysis of HUD and Treasury leveraging data.
[A] As a result of the data limitations described in this table, we
generally were unable to determine the extent to which agency-reported
leverage measures for the selected programs were over-or
underestimated.
[B] According to HUD officials, the agency has no mechanism to
determine whether zeros were nonresponses or $0 responses. Because it
is not possible to distinguish between nonreponses and $0 responses, we
were unable to determine the reliability of the leveraging data for the
CDBG and HOME programs. Further, since HUD started collecting
leveraging data for the CDBG program in December of 2005, only about
half of all program administrators had reported relevant data to the
agency. To the extent projects for which data were not available
leveraged funds, the total amount of leveraging that occurred in the
CDBG program potentially would be underestimated. See app. I for a more
detailed discussion of our assessment of the reliability of the
leveraging data for these programs.
[C] Treasury's potential underestimation of match leverage in the
program affects its calculation of institutional leverage (which
comprises match leverage and debt leverage) and total program leverage.
See app. III for a more detailed discussion on how Treasury calculated
a leverage measure for the CDFI program.
[D] Treasury only tracks taxpayers' compliance with rules for claiming
Low-Income Housing Tax Credits. No agency or organization collects data
that could be used to calculate leverage measures for the program.
[E] See 26 C.F.R. 1.45D-1(c). According to Treasury officials, CDEs
generally contribute more than the required minimum amount to qualified
low-income community investments; however, data were not available to
determine actual contributions.
[End of table]
In our assessment of HUD's and Treasury's use of leverage measures in
strategic planning, annual performance and budget documents, on their
Web sites, and in other published reports, we found that the agencies
did not routinely disclose the limitations to the leveraging data
(outlined in table 1) they used to compute leverage measures for the
selected programs. For example, the only place in which Treasury
included discussions of known limitations to the data used to calculate
a leverage measure for the CDFI program was in periodic agency reports
on the extent of leveraging in the program.[Footnote 32] Treasury's Web
site and key performance and budgeting documents provide little to no
information on data limitations associated with the CDFI program
leverage measure. Similarly, while HUD cited leverage measures on its
Web site and in budget documents for the HOME and HOPE VI programs,
respectively, the agency did not disclose the limitations of the data
used to compute the reported leverage measures for the
programs.[Footnote 33]
Further, no agency-specific or governmentwide guidance directs what
agencies should disclose about the leverage measures they report for
the selected programs; however, we regularly have reported on the need
for agencies to collect and report on credible and reliable data for
performance budgeting and other purposes.[Footnote 34] For example,
cautioning decision makers and others about significant data
limitations allows them to judge the credibility of the data and use
them in appropriate ways. We also noted that all data have limitations
that may hinder their use for certain purposes, and decision makers and
others may not have enough familiarity with the data to recognize the
significance of the shortcomings. Therefore, we concluded that
appropriate use of data may be fostered by clearly communicating how
and to what extent data limitations affect assessments of
performance.[Footnote 35] OMB also has stressed the importance of
making clear to policymakers and others what individual performance
indicators measure. According to OMB, doing so helps decision makers
understand what should be expected of an overall program.[Footnote 36]
To the extent that HUD and Treasury were not clear about the
limitations of the measures they calculated for the selected programs,
they potentially misrepresented (either positively or negatively) the
extent of leveraging that occurred in these programs. If decision
makers are unaware of the limitations of the agencies' reported
leverage measures and take them at face value, they could misuse them
in making funding decisions or performance evaluations on the programs
(which also may have implications on the budget process).
Agencies Also Generally Did Not Disclose Methods Used to Calculate
Leverage Measures Even Though Alternative Calculation Methods Can
Provide Significantly Different Results:
In our assessment of HUD's and Treasury's use of leverage measures in
strategic planning, annual performance and budget documents, on their
Web sites, and in other published reports, we also found that the
agencies did not routinely disclose information on the methods they
used to calculate leverage measures for the selected programs. For
instance, in its fiscal year 2008 budget justification, HUD reported
that the HOPE VI program leveraged $634 million over a 6-month period
in 2007, without further explanation of how the measure was derived.
Similarly, Treasury's Web site noted that on average CDFIs leveraged
program funds 20 to 1, but did not explain what types of funds (public
or private) were leveraged.
Based on our discussions with agency officials, we found that the
leverage measures HUD and Treasury calculated for each of the selected
programs generally described the ratio of all other funds contributed
to a program (including other federal, state, local, and private funds)
to program funds.[Footnote 37] However, these measures can be
calculated in multiple ways that describe leveraging from different
perspectives, such as the extent that federal funds are used with
nonfederal funds or public funds are used with private funds, which
underscore the importance of disclosing the calculation methods used.
As illustrated in figure 1, leverage measures for a single program
could vary considerably depending on how funding categories were
combined (that is, program funds, other federal funds, state and local
funds, and private funds).[Footnote 38]
Figure 1: Calculation Scenarios for a Hypothetical Program:
This figure is a bar graph showing calculation scenarios for a
hypothetical program.
[See PDF for image]
Source: GAO.
[End of figure]
Scenario A in figure 1 generally represents how the agencies presented
leverage measures for the selected programs. The alternate leverage
measures presented in scenarios B and C provide additional information
that could be more useful to policymakers and investors than measures
that describe the ratio of all other funds to program funds. For
example, to help inform decisions made as part of the annual
appropriations process, policymakers may be interested in determining
the extent of total federal contributions made to projects funded under
a particular program (scenario B). Alternatively, to assess the
potential risk of investing in a federally sponsored development
project in a low-income community, a private investor might be
interested in knowing the proportion of private investment to public
investment in the program (scenario C). Some private investors might
perceive a relatively low ratio as an indication that the program
carried a high level of investment risk and thus a higher potential for
losses.[Footnote 39] Further, more detailed information on all the
different sources of funding could be useful in describing the extent
to which one federal program is leveraging funds from another federal
program (that is, the extent to which federal programs cross-subsidize
one another) and could be particularly relevant to policymakers during
annual budget deliberations.
In addition, for the CDFI and New Markets Tax Credit programs (which
leverage at both the institutional and the project levels), disclosing
information on institutional and project-level leveraging could be more
useful to policymakers and investors than a total program leverage
measure.[Footnote 40] For example, providing such information would
assist policymakers and investors in understanding the extent to which
institutional leveraging could be used to manage project-level
investment risks--a program with a high institutional leverage ratio
but a low project leverage ratio might be one which invests in riskier
projects than a program with a low institutional leverage ratio but a
high project leverage ratio. As we discussed in our previous report on
leveraging federal funds, investments at the institutional level
generally are isolated from the investment risks associated with
discrete projects.[Footnote 41]
In appendix II we present multiple calculation scenarios for each of
the selected programs. Consistent with our hypothetical demonstrations
in figure 1, our calculations show considerably different results
between the leverage measures the agencies reported (that is, the ratio
of all other funds to program funds) and measures that describe either
(1) the ratios of nonfederal funds to federal funds and private funds
to public funds or (2) institutional and project leverage
ratios.[Footnote 42]
As a result of not having more specific information about how these
measures were calculated, decision makers would not have sufficient
information to understand their meaning and how they can and should be
used in performance assessment, budgeting, and other contexts. Further,
as previously discussed, there is no agency-specific or governmentwide
guidance on what agencies should report about how (or the extent to
which) leveraging occurs in their programs. However, in other contexts,
our prior work and that of OMB has stressed the value in agencies'
disclosing this type of information to ensure decision makers not only
are aware of what is being reported about a program, but how that
information can and should be used to inform their budget, performance,
and other decisions.[Footnote 43]
Leverage Measures Provide Basic Financial Information, but the Extent
to Which They Are Relevant for Assessing Program Performance Varies:
Leverage measures can provide basic financial information about the
programs included in our review; however, their relevance in assessing
the performance of these programs varies considerably. For all of the
programs we reviewed, leverage measures can describe inputs, or the
resources used to support program activities, and may be useful for
conveying basic financial information. To the extent that leveraging is
a goal or core (expected) activity of a program (as in the three
Treasury programs), leverage measures generally can describe program
outputs, or the products or services delivered (such as total leveraged
funds), and may be used along with other performance indicators to
assess the efficiency and effectiveness of a program in meeting its
goals. In cases where leveraging is not clearly and appropriately
linked to program goals and activities (as in the three HUD programs),
use of such measures to describe program outputs could be misleading
and result in adverse consequences, such as giving funding priority to
projects that leverage more over those that leverage less, but which
may fill a greater or more immediate need within a community. Although
leveraging had limited relevance to the goals and activities of the
selected HUD programs, we found that OMB and the agency often cited
leverage measures for the programs in performance-and budget-related
reviews and documents. Their continued use of leverage measures in
these contexts could unnecessarily encourage HUD to place more
importance on leveraging than meeting the stated goals of the CDBG,
HOME, and HOPE VI programs.
Leverage Measures Can Provide Basic Financial Information about a
Program, and if Linked to Program Goals and Core Activities, More
Detailed Performance Information:
Leverage measures generally can be used to describe the sources and
amounts of funds contributed to a program, and if linked to a program's
goals and core activities, they also can provide more detailed
information about the program's performance. On a basic level and for
all of the programs we reviewed, leverage measures convey information
on inputs--that is, the specific sources of funds used to implement
program activities. For example, leverage measures can provide
information on the relative contributions made by different types of
investors (private and public) to a program or project and the overall
resources committed--this information could be used to inform agency
budgeting exercises or financial analyses. To the extent that
leveraging is a goal or core (expected) activity of a program, leverage
measures generally can describe program outputs (in addition to program
inputs) and be used with other performance indicators to measure the
efficiency or effectiveness of a program in reaching its goals (see
fig. 2). Previously we have reported that for performance measures to
be useful in assessing program performance, they should be linked or
aligned with program goals and cover the activities that an entity is
expected to perform to support the intent of the program.[Footnote 44]
Generally, leveraging would not be an outcome measure for any of the
selected programs--outcomes describe program benefits or consequences
(such as the impact of leveraging on community development), whereas
outputs generally measure quantities produced (total dollars
leveraged).
Figure 2: Performance Measurement Model:
This figure is a flowchart showing the performance measurement model.
[See PDF for image]
Source: Adapted from OMB.
[End of figure]
The Importance of Leverage Measures in Assessing the Performance of the
Selected Programs Varies:
Leverage measures can be used to assess the performance of programs
that were designed to leverage (that is, in which leveraging is
directly related to the goals and core activities of the program), but
are less meaningful in assessing the performance programs that do not
have explicit leverage requirements. Each of the three Treasury
programs was designed to leverage other funds in a number of ways and,
as a result, leveraging directly relates to each program's goals and
core activities and leverage measures can be used to describe program
outputs. Under the CDFI program, CDFIs must match federal program funds
at least dollar-for-dollar with nonfederal funds as a condition of
receiving program funds. The match requirement is intended to increase
the sustainability of CDFIs (by increasing private-sector investment in
them) as well as their ability to make investments serving low-income
individuals and communities. Although not required to do so, CDFIs use
program and match funds to leverage debt and further increase their
lending resources. Funding recipients (for example, small businesses)
also may use their grants or loans from CDFIs to leverage additional
funds to help finance their projects. In this way, leveraging at the
project level also relates closely to the CDFI program's goal of
increasing investment in low-income individuals and communities.
Similarly, the tax credit programs were designed to automatically
generate private-sector equity investments in the production of
affordable housing (in the case of the Low-Income Housing Tax Credit
program) and community and economic development (in the case of the New
Markets Tax Credit program).[Footnote 45] Further, the application
processes for both programs were designed to encourage additional
leveraging. Under the Low-Income Housing Tax Credit program, in order
to limit the federal share of housing development project costs, states
are to provide no more tax credits to projects than necessary for their
financial viability.[Footnote 46] Under the New Markets Tax Credit
program, Treasury considers CDEs' potential to leverage other sources
of funds (in addition to the qualified low-income community investment
they plan to make using the tax credit equity) for the projects they
sponsor as a factor in scoring the tax credit allocation
applications.[Footnote 47]
In cases where leverage measures are not clearly and appropriately
linked to program goals and core activities, use of such measures to
describe program outputs could result in adverse consequences; for
example, by encouraging agencies to place more importance on leveraging
than on meeting their stated goals. This trade-off is directly apparent
in the use of leverage measures as outputs for the CDBG and HOME
programs.[Footnote 48] While leveraging may be a strategy some funding
recipients employ (either by choice or out of necessity) to meet these
programs' goals, none of these programs originally was designed to
leverage (meaning, leveraging generally is not a goal or core activity
in these programs). Thus, using leveraging to assess the success or
impact of these programs in meeting their goals may result in agencies
and funding recipients serving fewer lower-income communities or
households (as originally intended by these programs) and more moderate-
income communities and households (those that are better able to
attract additional funds because they pose relatively less risk to
investors).
HUD set a leveraging goal for the HOPE VI program in the agency's most
recent strategic plan and its fiscal year 2007 annual performance plan
and fiscal year 2008 budget justification. According to HUD officials,
while leveraging has long been a rating factor in the program's
application process, its relative importance in financing HOPE VI
developments has increased over time as program appropriations have
declined. While leveraging may help HUD meet the HOPE VI program goal
to create mixed-income communities, its use may involve trade-offs, as
it may conflict with another program goal--providing housing for
extremely-low, very-low, and low-income households.[Footnote 49] For
example, increased reliance on leveraged funds from other programs or
sources that may have different requirements (such as higher income
limits) potentially could affect the demographic composition of HOPE VI
developments.
Previously, we have reported several limitations to the usefulness of
leverage measures in providing detailed information about federal
programs and the projects they fund (regardless of whether or not those
programs were designed to leverage).[Footnote 50] Although leveraging
can be a useful tool and public-and private-sector officials regard it
favorably, according to many of the officials we contacted, if
considered independently of other information, leverage measures can
provide misleading information about the success or impact of a program
or project. For example, many said that factors such as the local
economy or availability of investors within a certain geographic area
could have a positive or negative impact on a project's ability to
leverage additional funds, and thus its leverage ratio. That is,
projects in vibrant communities likely may have higher leverage ratios
than those in distressed communities. As a result, leverage measures
are not sufficient to make judgments about the relative success of
projects or programs without other descriptive information. Leverage
measures also do not account for the level of substitution of federal
funds for otherwise available private funds that might occur in
programs or projects. Although difficult to measure, information on
substitution might be useful in assessing how effectively federal funds
were utilized in a program or project. Officials we contacted noted
that having information on the risk position of different contributions
to a project might be useful in assessing the extent of substitution
that occurred. For instance, the level of substitution in a project in
which the federal government assumed more risk (by taking a subordinate
position) than nonfederal investors could be lower than the level of
substitution in a project in which the federal government assumed less
risk (by taking a senior position).[Footnote 51]
OMB and the Agencies Did Not Always Link Leveraging to Program Goals
and Core Activities in Performance-related Reviews and Reports:
When OMB and the agencies cited leverage measures in performance-and
budget-related reviews and documents, they did not always link
leveraging to program goals and core activities--in some cases, OMB and
the agencies used leveraging to assess the performance of the selected
programs despite its limited relevance to program goals and core
activities. According to OMB officials, the agency considers leveraging
to be an output measure for each of the selected HUD programs.
Consistent with this view, OMB used leveraging as an output measure in
its PART reviews of these programs, although leveraging generally was
not linked to the goals and core activities of the programs.[Footnote
52] For example, in its 2003 PART review of the CDBG program, OMB
recommended that HUD implement a new performance measurement system
that included information on the amount of money leveraged from other
sources.[Footnote 53] The agency developed steps to address this
recommendation in the program improvement plan it developed with OMB in
2006 (in response to the PART assessment's finding that the program
lacked specific annual performance measures that demonstrated progress
on achieving long-term goals).[Footnote 54] We have noted that federal
programs, in particular federal block grant programs, have faced
difficulties but could benefit from defining program goals and
performance measures that go beyond describing program activities to
describe outcomes or results.[Footnote 55] However, because leveraging
is not a required activity or explicit goal of the CDBG program (as
discussed previously), its value in evaluating the performance of the
program is limited. Further, in its PART review of the HOME program,
OMB used leverage measures to compare the performance of the HOME
program with that of the CDBG program. Such a comparison does not
facilitate evaluations of these programs in the context of their
intended goals (neither of which is to leverage).
While using leveraging as an output measure for the CDFI and New
Markets Tax Credit programs is consistent with the programs' goals and
core activities as discussed above, OMB identified leveraging as an
outcome measure for the CDFI program in its 2004 PART review despite
the fact that, as discussed previously, leveraging cannot be used to
measure the impact of the program.[Footnote 56] Further, the agency
equated leveraging with program effectiveness in its 2004 PART review
of the New Markets Tax Credit program. (As described in fig. 2, outcome
measures are used to assess the effectiveness of programs in achieving
desired results. As we have discussed throughout this report, outcome
measures should be designed to assess the benefits or consequences of a
program--leverage measures by themselves cannot provide information on
the impact these programs have had on their targeted populations and
communities.)
As we observed in our 2004 review of OMB's PART process, the goals and
measures OMB defines in its PART reviews are designed to meet the needs
of executive decision makers during the budget formulation process, and
thus may be inconsistent with the goals and measures federal agencies
have developed in response to GPRA, which may be developed at a higher,
strategic level and less relevant to OMB's budget decision-making
process.[Footnote 57] As a result of OMB's focus on the budget process,
we found that its judgment about appropriate goals and measures for a
program may be substituted for agency judgments. These findings
generally are consistent with our observations on OMB's use of leverage
measures in the PART reviews of the selected programs we reviewed for
this report. We observed that the agencies identified leveraging as a
performance measure in their performance-and budget-related reports for
some of the selected programs despite its sometimes limited relevance
to program goals and core activities.
Table 2 describes HUD's use of leverage measures for the HOME and HOPE
VI programs in its strategic planning and other performance-and budget-
related documents or contexts.[Footnote 58] In the case of the HOME
program, although leveraging was not linked to the program's goals and
core activities, HUD equated more leveraging with better performance by
ranking states and localities on their ability to leverage other
sources of funds. For the HOPE VI program, HUD primarily used
leveraging as a measure for its goal of providing decent, affordable
housing through the improvement of the physical quality of public
housing. However, HUD generally did not discuss how leveraging would
help the agency in achieving this goal. HUD also linked leveraging to
the HOPE VI goal of creating mixed-income housing. Although increased
leveraging in a program designed to provide affordable housing could
result in trade-offs, HUD's performance-and budget-related documents
did not discuss the impact of (or the potential unintended consequences
of) leveraging on the ability of the program to meet this goal.
Table 2: HUD's Use of Leverage Measures in Performance Assessment and
Other Documents:
Program and document: CDBG: None;
Measure and linkage to program goals: Not applicable.
Program and document: HOME: Performance Snapshot Reports;
Measure and linkage to program goals: Measure: According to HUD's
explanation of performance categories presented in the snapshot
reports, a leveraging ratio of 4 to 1 is considered indicative of
significant leveraging. Therefore, any state or locality with a
leveraging ratio of 4 to 1 and greater would receive a designation of
100 percent (a ranking of 1). Any state or locality with a leveraging
ratio of less than 4 to 1 would receive a lower score. For example, a
leveraging ratio of 2 to 1 (half of 4 to 1) would receive a designation
of 50 percent;
Linkage: Although used to make performance-related comparisons among
grantees, HUD does not link leveraging to HOME program goals in these
reports.
Program and document: HOPE VI: 2006-2011 Strategic Plan;
Measure and linkage to program goals: The HOPE VI program will leverage
$4 billion in private financing between 2006 and 2011;
Linkage: Leveraging in the HOPE VI program is linked to HUD's mission
to promote decent, affordable housing. Programmatic strategic goals
under this mission include, among other things, (1) expanding access to
and availability of decent, affordable rental housing; (2) improving
the physical quality of public and assisted housing; and (3)
facilitating more effective delivery of affordable housing by reforming
public housing. However, the plan does not provide details on how
leveraging in the HOPE VI program facilitates HUD meeting the mission
and related strategic goals.
Program and document: HOPE VI: Fiscal Year 2007 Annual Performance
Plan;
Measure and linkage to program goals: Measure: The HOPE VI program will
leverage $800 million in other financing in fiscal year 2007;
Linkage: Leveraging in the HOPE VI program is linked to HUD's mission
to promote decent, affordable housing and more specifically to the
agency's strategic goal of improving the physical quality of public
housing. However, in HUD's more detailed discussion of how leveraging
would help achieve this mission and its related goal, the agency links
leveraging to the creation of mixed-income communities, rather than the
stated goal (improving the physical quality of public housing).
Program and document: HOPE VI: Fiscal Year 2008 Budget Justification;
Measure and linkage to program goals: Measure: The HOPE VI program will
leverage $800 million in other financing in fiscal year 2008;
Linkage: Leveraging is linked to the creation of mixed-income
communities. HUD specifically asserts that the formation of new public
and private partnerships is key in ensuring the long-term
sustainability of public housing development and the leveraging of
public and private resources to transform isolated public housing
communities into sustainable, mixed-income communities with a wide
range of family incomes. HUD provides no discussion of how leveraging
links to or positively or negatively affects another of the program's
missions--to promote decent, affordable housing--or strategic goals--to
improve the physical quality of public housing.
Source: HUD publications.
Note: According to HUD officials, the $800 million HOPE VI leverage
goal reported in the agency's fiscal year 2007 annual performance plan
and fiscal year 2008 budget justification recently was revised to $650
million. In comments on a draft of this report, HUD officials noted
that the agency posted revised performance documents on its Web site
reflecting this change.
[End of table]
Finally, as described in table 3, Treasury generally linked leveraging
with the goals and core activities of the CDFI and New Markets Tax
Credit programs.[Footnote 59] For example, Treasury noted that
leveraging in the CDFI program helps build CDFIs' capacity to make
loans and other investments in low-income communities. Because Treasury
to date has not reported publicly the extent of leveraging in the New
Markets Tax Credit program, the agency's performance-and budget-related
documents only discuss the extent of institutional leverage in the
program. As with the CDFI program, Treasury linked institutional
leveraging to the program's goal of attracting private-sector capital
to low-income communities.
Table 3: Treasury's Use of Leverage Measures in Performance Assessment
and Other Documents:
Program and document: CDFI: Fiscal Year 2006 Performance and
Accountability Report;
Measure and linkage to program goals: Measure: 186 CDFIs leveraged $1.4
billion in fiscal year 2005;
Linkage: Total leveraging is used to measure progress in meeting the
program's goal to build the capacity and coverage of CDFIs to provide
credit, capital, and related services to otherwise underserved markets.
According to the report, Treasury provides financial assistance through
the CDFI program in the form of grants, loans, and equity investments
to CDFIs. Financial assistance awards are made to CDFIs that have
comprehensive business plans for creating community development impact
and that demonstrate an ability to leverage private-sector sources of
capital.
Program and document: CDFI: Fiscal Year 2006 Justification for
Appropriations and Performance Plans (CDFI Fund breakout);
Measure and linkage to program goals: Measure: The approximately $12.4
million in fiscal year 2008 program funds should result in an
additional $335 million raised and deployed in low-income communities;
Linkage: Same as above.
Program and document: Low-Income Housing Tax Credit: None;
Measure and linkage to program goals: Not applicable.
Program and document: New Markets Tax Credit: Fiscal Year 2006
Performance and Accountability Report;
Measure and linkage to program goals: Measure: Treasury awarded $4.1
billion dollars to 63 CDEs in fiscal year 2006 (institutional
leverage);
Linkage: Equity investments (or funds leveraged at the institutional
level) are used to measure progress in meeting the program's goal to
attract private-sector capital into low-income communities through
CDEs. The New Markets Tax Credit program is intended to spur private-
sector capital into low-income areas through CDEs, which in turn make
loans and equity investments in businesses and real estate projects in
low-income communities. By making an equity investment in a CDE,
individual and corporate investors can receive a tax credit against
their federal income taxes worth 39 percent of the value of the amount
invested in the CDE over 7 years.
Program and document: New Markets Tax Credit: Fiscal Year 2006
Justification for Appropriations and Performance Plans (CDFI Fund
breakout);
Measure and linkage to program goals: Measure: The fiscal year 2008
allocation round will provide tax credit allocations supporting $3.5
billion in investor capital (institutional leverage);
Linkage: Same as above.
Source: Treasury publications.
[End of table]
Conclusions:
With the increased focus of federal agencies on performance management,
budgeting, and financial reporting, leveraging has come to be seen as
an effective and efficient means of delivering more impact per dollar
of federal investment, particularly in a period of increasingly tight
budgets and competing funding priorities. While agencies have collected
and presented leveraging information in strategic planning,
performance, and budget reports, and on their Web sites, agencies
disclose little or no information on methods of data collection or how
leverage measures were calculated, in part because there is no agency-
specific or governmentwide guidance on how to calculate, describe, and
use leverage measures in a manner that is consistent with the programs'
design. Information on methodology is important in the leveraging
context because of the limitations of leveraging measures and data
collection issues. For example, in the case of the CDBG and HOME
programs, leveraging may be underestimated because HUD's database does
not distinguish between zero responses (for example, where no
leveraging occurred) and blank responses (for example, where leveraging
data may be incomplete). Moreover, measures such as ratios may not
disclose the details necessary to understand which component funding
sources were being compared, and as demonstrated, the ratios can vary
considerably depending on what information an agency is trying to
convey about a program (for example, the extent of public or private
investment in a program). Further, data collection and completeness are
issues because not all the programs are required to report leveraging,
and in many cases agencies are unable to capture data on all leveraging
that may be occurring in a program (for example, project leveraging).
Absent specific information on how leverage measures were calculated
and their potential limitations, decision makers do not have sufficient
information to understand their meaning and how they can and should be
used in performance assessment, budgeting, and other contexts.
Moreover, the relevance of leveraging to performance measurement is
dependent on the context of the program being analyzed. Because
leveraging is not an intended activity carried out to achieve program
goals or a goal unto itself for some of the selected housing and
community and economic development programs in our review, measures
such as ratios are not indicative of program or project performance
(outcomes and impact). Rather, such measures are indicative only of
resource utilization. Nevertheless, even in cases where they were not
reflective of program performance, agencies presented leverage measures
in strategic plans, annual performance plans, performance and
accountability reports, and budget justifications. The use of leverage
measures in such contexts could lead decision makers to presuppose that
the information was indicative of program impacts in cases where
leveraging actually might say very little about the success of a
program, such as the ability of a program to improve the living
conditions of the urban poor.
Despite the issues surrounding the utility of leverage measures, we
note the valid and useful purposes for which the measures may be used,
particularly in instances where leveraging is an intended activity or
goal. For instance, decision makers and practitioners in the area of
affordable housing and community and economic development may utilize
leverage measures to report basic information on how federal funds were
combined with other funds for a program or project. Such information
could be instructive in ascertaining trends in the involvement of
private-sector investors or local governments in federally sponsored
initiatives, or identifying demographic trends that could adversely or
positively affect the ability of program funds to attract other funds.
Additionally, the measures may aid management and Congress in their
oversight of programs and strategic planning for future budgets.
Further, when directly linked to program goals and activities and
considered with other performance measures, leverage measures also
could provide insight into the success of a program, including its
impact on targeted populations and communities.
The valid and useful purposes to which leverage measures may be put
underline the importance of transparency for federal agencies in
communicating the limitations of such measures and how they are
calculated. The agencies administering the housing and community and
economic development programs we reviewed could improve the
transparency of the leverage measures they use by including information
about the completeness and accuracy of the data and methods used to
compute the measures. Further, the agencies could discuss the relevance
of leveraging to a program's stated goals and activities. The
publication of such information in conjunction with the measures
themselves would increase the accuracy of the information being
conveyed and provide perspectives that would allow various users to
assess the potential of the measures to serve as relevant and accurate
indicators of program or project outputs and, in some cases, outcomes
or impact. However, the opportunities to better describe, assess, and
report the role of leveraging in housing and community and economic
development programs do not rest solely with the agencies administering
those programs. OMB, because it plays a key role in assessing the
performance of federal agencies and developing and tracking compliance
with performance goals, has an opportunity to refine its understanding
and use of leverage measures in future PART and other performance
reviews by carefully considering the role of leveraging in carrying out
program goals and activities. Specifically, in its performance
assessments of the selected programs, OMB could provide information on
how leveraging may support or conflict with a program's intended
purpose. This is particularly important because the accuracy of
measures and the relationship of leveraging to program goals and thus
performance can vary considerably across the housing and community and
economic development programs we reviewed.
Recommendations for Executive Action:
To ensure that leverage measures provide accurate, useful, and relevant
information to Congress and others, we recommend that the Secretaries
of HUD and the Treasury consider disclosing the following when they
publish such measures for the programs included in our review:
* Presentation of leverage measures should be accompanied by
information about the completeness and accuracy of the data and the
method(s) used to calculate the measures (for example, with leverage
ratios, information on what sources of funds were compared, such as
private funds to public funds or nonfederal funds to federal funds).
* Presentation of leverage measures should be accompanied by a
discussion of the relevance of the measure in assessing the program's
performance. For example, the agencies should discuss the extent to
which leverage measures are linked to program goals and core
activities.
We further recommend that the Director of OMB:
* provide guidance to help agencies determine how to calculate,
describe, and use leverage measures in a manner consistent with the
programs' design; and:
* re-evaluate the use of leverage measures and disclose their relevance
to program goals and activities in future PART or other performance
reviews of the selected programs.
Agency Comments and Our Evaluation:
We received written comments on a draft of this report from HUD and
Treasury, which are included in appendixes V and VI, respectively.
HUD's Office of Public and Indian Housing also provided technical
comments related to the HOPE VI program, which we incorporated as
appropriate. We also provided a draft of this report to OMB for review,
but no comments were provided.
In a letter from the Acting Deputy Assistant Secretary for Grant
Programs, HUD noted that it was pleased with the results presented in
our draft report, but provided several detailed comments on and
suggested changes to our findings related to the CDBG and HOME programs
(see app. V). Specifically, HUD expressed concern that the draft report
(1) did not sufficiently emphasize that the CDBG and HOME programs do
not have statutory or regulatory leveraging requirements; (2) did not
sufficiently emphasize that the agency currently does not publish a
leverage measure for the CDBG program; (3) incorrectly stated that the
agency did not disclose limitations to the data or methods used to
calculate leverage measures for the HOME program, which are reported on
HUD's Web site; and (4) contends that leveraging affects the funding
decisions HUD makes for CDBG and HOME (HUD noted that all funding
decisions are made at the state or local level and are not approved by
the agency).
With respect to HUD's first two concerns, we incorporated additional
language into the report to further emphasize that the CDBG and HOME
programs do not have leveraging requirements and that the agency does
not publish a leverage measure for the CDBG program. In its letter, HUD
agreed to work to improve the quality of leveraging data CDBG grantees
report to the agency, which would address, in part, our recommendation
that the agencies disclose information about the completeness and
accuracy of the data and the method(s) used to calculate leverage
measures.
Concerning HUD's comment that the draft report incorrectly stated that
the agency did not disclose limitations to the data or methods used to
calculate leverage measures for the HOME program, which are reported on
HUD's Web site, we acknowledge that HUD's Web site included information
on the method used to calculate leverage measures for the HOME program
(that is, the ratio of other funds to program funds). However, HUD has
not provided information on the limitations to the data used to
calculate those measures. Specifically, the database HUD used to
collect leveraging data for the program did not distinguish between
nonresponses, which default to zero, and actual entries of zero;
assuming that some grantees failed to enter funding information, the
total amount of leveraging that occurred in the program (or in a
specific state or locality) potentially would be underestimated.
Accordingly, we did not change the report.
Finally, with respect to HUD's concern that the draft report contends
that leveraging affects the funding decisions HUD makes for CDBG and
HOME, our report did not state that HUD or grantees make funding
decisions based on leveraging; rather, the report noted the potential
consequences of using leveraging as a performance indicator for
programs that were not designed to leverage. Specifically, we found
that leveraging may be a strategy some funding recipients employ,
either by choice or out of necessity, to meet the goals of the CDBG and
HOME programs. Thus, using a leverage measure to assess the impact or
success in meeting goals may create adverse or conflicting incentives
for the agency and its grantees as well as Congress and other decision
makers; for example, by giving funding priority to projects that
leverage more over those that leverage less, but which may fill a
greater or more immediate need within a community. In response to this
comment, we added language to the report to emphasize that HUD has not
identified leveraging as a performance measure for either program.
In a letter from the Director of the Community Development Financial
Institutions Fund, Treasury expressed appreciation for our finding that
each of the agency's programs included in our review was designed to
leverage. Although Treasury did not specifically comment on our
recommendations, it provided several detailed comments primarily
related to the agency's calculation of leverage measures for the CDFI
and New Markets Tax Credit programs (see app. VI). Specifically,
Treasury commented on our findings that (1) the leverage measures
Treasury reported for its programs lacked transparency because the
agency did not disclose the limitations of the data or the methods used
to calculate them; (2) the leverage measures did not reflect the actual
extent of leveraging in the CDFI program due to incomplete data; (3)
missing project-level data for the New Markets Tax Credit program
potentially led to misestimations of leveraging in the program; and (4)
the leverage measure Treasury calculated for the New Markets Tax
program was a multiplier ratio, not a leverage ratio.
With respect to its first comment on our findings, Treasury stated that
on multiple occasions the agency has publicly disclosed its calculation
method for the CDFI program. In the report, we listed two publications
in which Treasury disclosed its calculation methodologies and
limitations to the data it used to compute a leverage ratio for the
CDFI program.[Footnote 60] To this list, we added the additional report
Treasury cited in its letter.[Footnote 61] However, we continue to
believe that disclosure of the methodologies and limitations of the
data used to calculate the leverage measures is important, particularly
in key budget and performance documents, which policymakers often rely
on to make funding and management decisions. As discussed in the
report, Treasury did not disclose such information about its leverage
calculation for the CDFI program in these key documents. For example,
in the fiscal year 2006 Performance and Accountability Report and
Justification for Appropriations and Performance Plans, Treasury
reported leverage ratios for the CDFI program and emphasized its
importance in achieving program goals, but did not include any
discussions of the measures' data limitations or calculation methods.
Accordingly, we did not change our finding that Treasury's reporting of
such information was inconsistent and that it should further disclose
its data limitations and calculation methods in key budget and
performance documents.
Concerning Treasury's comment on our finding that the leverage measure
the agency calculated for the CDFI program did not reflect the actual
extent of leveraging due to incomplete data, Treasury stated that
although it was aware that the match leverage--that is, the ratio of
nonfederal match funds to program funds--may actually exceed the
statutory requirement of a 1 to 1 ratio, it is not appropriate or
necessary to include excess matching funds that exceed the requirement.
We continue to believe that excluding excess matching funds from the
leverage calculation (which typically includes all other sources of
funds) understates the actual extent of leveraging that occurs in the
program. Accordingly, we did not change the report in this regard. If
Treasury chooses to continue to exclude such amounts from future,
published leverage calculations for the program, we believe that it
should disclose this and its potential impact on the leverage measure,
consistent with the recommendations included in this report.
Concerning Treasury's comment on our finding that missing project-level
data for the New Markets Tax Credit program potentially led to
misestimations of leveraging in the program, Treasury stated in its
letter that the leverage measure for the program would not
substantially be different if complete data were available and that the
calculated measures provided a reasonable approximation of the
leveraging that occurs in the program. We reported that (1) leverage
data were not available for 26 percent of New Markets Tax Credit
projects and (2) Treasury assumed that CDEs contribute 100 percent of
tax credit equity to qualified low-income community investments, even
though CDEs are permitted to retain up to 15 percent of such equity for
administrative and other purposes. We noted that the former case could
lead to an underestimation of the extent of leveraging and the latter
an overestimation of the extent of leveraging that occurred in the
program. As discussed above with respect to the CDFI program, these
data limitations potentially could have an impact on the leverage
measure Treasury calculated for the program. In its letter, Treasury
agreed with our description of these limitations, but did not provide
any specific evidence of the impact of missing project-level data on
these measures. Treasury also acknowledged the importance of disclosing
such information, stating it would make every effort to include a
discussion of these and other data limitations, as well as its
calculation methodologies, when and if it publishes leverage measure
for the program.[Footnote 62] In response to these comments, we did not
change the report.
Finally, with respect to our finding that the leverage measure Treasury
calculated for the New Markets Tax Credit program for purposes of this
report was a multiplier ratio, Treasury stated that the measure was a
leverage ratio, calculated consistent with GAO guidance outlined in the
report. However, we reported the measure Treasury reported for the New
Markets Tax Credit program was not a leverage ratio, but rather a money
multiplier or multiplier ratio. A multiplier ratio measures the total
amount of investment $1 in tax credits potentially can generate in low-
income communities, whereas a leverage ratio measures the additional
amount of investment relative to a source of funds (such as program
funds). According to Treasury officials with whom we spoke, the agency
included the cost of the credit ($0.25) on "both sides of the ratio,"
consistent with the calculation of a multiplier ratio, but overstating
the extent of leveraging that occurred in the program. Our purpose in
making a distinction between leverage ratios and multiplier ratios was
to highlight the need for adequate disclosure of calculation methods
and data limitations so that decision makers understand how to
interpret these measures and how these measures compare with those
reported by other programs. Without such information, it would not be
possible for decision makers to assess the reliability of the measures
or the comparability of the measures reported by other programs. If
Treasury publishes the measure it calculated for the New Markets Tax
Credit program, we believe it is incumbent upon the agency to provide a
discussion as to how the measure was calculated in an attempt to
provide complete information to decision makers (see app. III).
Further, Treasury acknowledged in its letter it would do so, stating it
would "make every effort to include a discussion of data methodologies
and limitations" when it publishes leverage measure for the program.
Accordingly, we did not change the report in response to this comment.
HUD's and Treasury's letters also included several comments that were
technical in nature, which we incorporated as appropriate.
As agreed with your office, unless you publicly announce the contents
of this report earlier, we plan no further distribution until 30 days
from the report date. At that time, we will send copies of this report
to the Ranking Member, Subcommittee on Housing and Community
Opportunity, Committee on Financial Services, the Secretaries of
Housing and Urban Development and the Department of the Treasury, the
Director of the Office of Management and Budget, and other interested
congressional committees. We also will make copies available to others
upon request. In addition, the report will be available at no charge on
the GAO Web site at [hyperlink, http://www.gao.gov].
If you or your staff have any questions concerning this report, please
contact me at (202) 512-8678 or at [email protected]. Contact points for
our Office of Congressional Relations and Public Affairs may be found
on the last page of this report. Key contributors to this report are
listed in appendix VII.
Sincerely,
Signed by:
William B. Shear:
Director, Financial Markets and Community Investment:
[End of section]
Appendix I: Objectives, Scope, and Methodology:
The objectives of this report were to examine (1) the leverage measures
the Department of Housing and Urban Development (HUD) and the
Department of the Treasury (Treasury) reported for the selected housing
and community and economic development programs and the transparency of
the data and methods used to calculate them and (2) the relevance of
leverage measures in assessing the performance of the selected
programs. Our review focused on HUD's Community Development Block Grant
(CDBG), HOME Investment Partnership (HOME), and HOPE VI programs and
the Treasury's Community Development Financial Institutions (CDFI), Low-
Income Housing Tax Credit, and New Markets Tax Credit programs.
To examine the leverage measures HUD and Treasury reported for each of
the selected programs and the transparency of the data and methods used
to calculate them, we reviewed relevant program regulations and
guidance, our prior reports, and reports of others, and interviewed
agency officials and other stakeholders. Based on this information, we
requested from HUD and Treasury data they use to measure the extent of
leveraging (for example, data on the sources and amounts of funds, or
other financial data, commonly referred to as "leveraging data") in the
CDBG, HOME, and HOPE VI programs and the CDFI and New Markets Tax
Credit programs, respectively. We did not request Low-Income Housing
Tax Credit data from HUD or Treasury because neither maintains a
database with detailed information on leveraging.[Footnote 63]
* For both the CDBG and HOME programs, we requested leveraging data on
completed program activities that were aggregated at the local level
from HUD's Integrated Disbursement and Information System (IDIS), which
contains information on activities funded by a number of grant programs
(including the CDBG and HOME programs).[Footnote 64] The CDBG data were
from December 1, 2005, and May 1, 2007, and the HOME data were from
October 1, 2005, and September 30, 2006.[Footnote 65] To assess the
reliability of the data for both programs we (1) performed basic
electronic testing of data elements associated with the financing used
by state and local agencies that administer the programs--for example,
we checked for missing data; (2) reviewed existing information about
the data and IDIS; (3) replicated the leverage measure that HUD
reported for each program; and (4) interviewed agency officials
knowledgeable about the data. As a result of these tests, we found
several limitations with these data, specifically that they were
largely self-reported by program administrators and were not validated.
In addition, IDIS does not distinguish between nonresponses, which
default to zero, and actual zero (that is, $0) responses; as such, the
data may underreport the total amount of leveraging that occurred in
the programs. Further, the data may be incomplete because HUD does not
require state and local agencies to report leveraging data because
leveraging is not a required activity in either the CDBG or the HOME
program, and HUD only started collecting leveraging data for the CDBG
program in December 2005 (only about half of all program administrators
have reported relevant data to the agency). Due to these limitations,
we were unable to determine the reliability of the precise dollar
amounts that were used in combination with the CDBG and HOME funds. We
use the leverage measures that HUD derived from the data to illustrate
how leverage measures can be calculated in different ways, but the
values should not be used to represent actual dollars leveraged.
* To assess the reliability of HUD's HOPE VI program leveraging data on
the 55 HOPE VI projects completed (that is, projects in which all
phases of construction were fully completed and actual funding amounts
were reported) as of March 2006, we (1) performed basic electronic
testing of data elements associated with the financing used by the
public housing agencies that administer the program; (2) reviewed
existing information about the data and HUD's HOPE VI Internet-based
Grant Management Reporting System Prototype (HOPE VI database); and (3)
interviewed HUD officials knowledgeable about the data.[Footnote 66] In
addition, we interviewed officials from five randomly selected public
housing agencies (PHA) that received a HOPE VI grant to determine the
accuracy and completeness of the data in the HOPE VI database as it
pertained to the PHAs' specific HOPE VI project. We determined that the
data were sufficiently reliable for the purpose of this
report.[Footnote 67]
* For the CDFI program, we discussed with agency officials the
calculation method used to compute the program's leverage measure,
including any assumptions made, the completeness and accuracy of the
data used in the calculation, and any other known limitations to the
measure or the data used to calculate it. Unlike the CDBG, HOME, and
HOPE VI programs, we did not request project-level data Treasury uses
to calculate a leverage measure for the program.[Footnote 68] Rather,
Treasury provided us with a spreadsheet containing the calculation
method and nationally aggregated data used to calculate leverage
measures for each of the last 6 reporting years. We determined that
Treasury's calculation method was appropriate and supporting data were
sufficiently reliable for the purpose of calculating an approximation
of the funds being leveraged in the program. However, based on our
conversations with agency officials, we also noted several limitations
in Treasury's calculation method and the supporting data. Specifically
(1) the data were largely self-reported by CDFIs and were not validated
and (2) Treasury assumed that matching contributions do not exceed $1
for every $1 in program funds, which likely understates the extent of
institutional-level leveraging in the program (because many of the
CDFIs exceed the match requirement, according to Treasury officials).
* To assess the reliability of the data Treasury provided on project-
level leveraging in the New Markets Tax Credit program, we (1)
performed basic electronic tests of the data elements associated with
the financing used by Community Development Entities (CDE), (2)
reviewed existing information about the data, (3) replicated the
project-level leverage measure Treasury calculated for the program, and
(4) interviewed agency officials knowledgeable about the data. We
determined that the project-level data were sufficiently reliable for
purposes of calculating a project-level leverage measure for the
program. However, we also noted some limitations in the project-level
data, specifically that (1) they were largely self-reported by CDEs and
were not validated, and (2) about 26 percent (139 out of 538) of the
CDEs that were awarded New Markets Tax Credits did not report data.
* To determine what leveraging data were available for the Low-Income
Housing Tax Credit program at the state level and whether such data
were maintained electronically, we conducted a telephone survey of the
entire population of 57 allocating agencies, which included 50 state
agencies, the District of Columbia, Puerto Rico, and the Virgin
Islands; one suballocating agency in the District of Columbia; two
suballocating agencies in the State of New York; and a suballocating
agency in Chicago.[Footnote 69] Our pretested survey achieved a 79
percent response rate. On the basis of 45 responses to the following
questions--(1) Does your agency have data in its database on the
specific types of financing sources that are used in each Low-Income
Housing Tax Credit project? and (2) Does your agency have the dollar
amounts contributed by each financing source used in the project in the
database?--we found that 25 allocating agencies collect the dollar
amounts contributed by specific financing sources and keep that data
electronically.[Footnote 70] Because not all allocating agencies
collected leveraging data and those that did used different software
applications to maintain their data, we determined that it would be
difficult to collect aggregate data to report a national leverage
measure for the program.
To examine the relevance of leverage measures in assessing the
performance of the selected programs, we reviewed our reports and those
of the Office of Management and Budget (OMB) on performance
measurement; agency strategic plans and annual performance plans,
budget justifications and performance and accountability reports; and
industry, and other literature such as agency press releases and Web
sites. We also interviewed representatives from Treasury, HUD, and OMB.
Additionally, we interviewed representatives of the following industry
groups and other organizations involved in housing and community and
economic development initiatives:
* City of Chicago Department of Housing;
* Coalition of Community Development Financial Institutions;
* Coastal Enterprises, Inc;
* Community First Fund;
* Council of State Community and Economic Development Agencies;
* Enterprise Community Partners;
* International Economic Development Council;
* Harvard University's Joint Center for Housing Studies;
* John D. and Catherine T. MacArthur Foundation;
* Living Cities: The National Community Development Initiative;
* Local Initiatives Support Corporation;
* National Association of Affordable Housing Lenders;
* National Association of Development Organizations;
* National Association of Housing and Redevelopment Officials;
* National Community Development Association;
* National Community Investment Fund;
* National Council of State Housing Agencies;
* National Development Council;
* National Urban League;
* NeighborWorks America;
* New Hampshire Community Loan Fund;
* New Markets Tax Credit Coalition;
* Reinvestment Fund; and:
* Western Massachusetts Enterprise Fund, Inc.
Further, as part of this work, we conducted site visits and collected
information on how federal funds have been leveraged in housing and
community and economic development projects, 20 of which we toured.
Specifically, we conducted site visits in Chicago, Illinois; Laredo and
San Antonio, Texas; Chester and Philadelphia, Pennsylvania; Portland
and Salem, Oregon; and Seattle and Tokeland, Washington. We selected
these areas to obtain perspectives from a variety of regions with
attributes such as difficult-to-develop areas, rural and urban
classifications, and lower-and higher-cost areas that affect the extent
of leveraging. We used a nongeneralizable, illustrative sampling
approach to select a range of housing and community and economic
development projects or initiatives to tour. More specifically, our
criteria were (1) projects were substantially completed in the last 5
years, (2) funds of each of the programs in our review were utilized in
at least one of the projects selected, and (3) projects had two or more
funding sources. In the cases where program officials maintained
comprehensive lists of projects, we used such lists to randomly select
projects; otherwise, we used available project information (for
example, from program administrator Web sites) in conjunction with
information from program administrators to select projects that
generally met our criteria. In selecting the projects we illustrate in
this report, we further considered several factors including the
availability and completeness of leveraging data, the creativity in the
projects' financing and design, the type of development or initiative,
whether the project was located in a rural area, and the general
geographic location of the project. Appendix II provides examples of
how federal funds have been leveraged in the selected programs.
We conducted this performance audit in Chicago, Illinois; San Antonio
and Laredo, Texas; Philadelphia and Chester, Pennsylvania; Portland and
Salem, Oregon; Seattle and Tokeland, Washington; and Washington, D.C.,
from November 2006 to January 2008 in accordance with generally
accepted government auditing standards. Those standards require that we
plan and perform the audit to obtain sufficient, appropriate evidence
to provide a reasonable basis for our findings and conclusions based on
our audit objectives. We believe that the evidence obtained provides a
reasonable basis for our findings and conclusions based on our audit
objectives.
[End of section]
Appendix II: Transactions Structures for the Selected Treasury
Programs:
Figures 3 through 6 describe how the CDFI, Low-Income Housing Tax
Credit, and New Markets Tax Credit programs leverage funds for housing
and community and economic development.[Footnote 71]
CDFI Program:
As illustrated in figure 3, CDFIs must match CDFI program funds dollar-
for-dollar with funds from nonfederal sources such as local governments
or private foundations (match leverage). CDFIs use these program and
match funds to attract private debt from lenders (debt leverage).
Together, match and debt leverage represent institutional leverage in
the CDFI program. CDFIs use the pooled equity and debt to make loans to
a number of development projects. Additional leverage also may occur at
the project level--individual projects may use their CDFI funds to
leverage funding (equity or debt) from other investors, such as
foundations, nonprofits, banks, and local governments. Projects
(borrowers) repay principal and interest to their investors, including
the CDFI. CDFIs use these payments to make subsequent loans to
additional projects and repay lenders.
Figure 3: CDFI Transaction Structure:
This figure is a flowchart showing CDFI transaction structure.
[See PDF for image]
Source: GAO.
[End of figure]
Low-Income Housing Tax Credit Program:
As illustrated in figure 4, states, through their state housing finance
agencies, are authorized to allocate Low-Income Housing Tax Credits to
housing projects. Project developers can sell their tax credits
directly to an investor(s) or a syndicator (which assembles a group of
investors and acts as the group's representative). The money investors
pay for the tax credits is paid into the projects as equity financing.
Generally, investors (including individuals, foundations, and state and
local governments) contribute this equity, which is combined with non-
tax credit financing sources (such as mortgages) in individual projects
to fund development costs.
Figure 4: Low-Income Housing Tax Credit Syndication Transaction
Structure:
This figure is a flowchart showing low-income housing tax credit
syndication transaction structure.
[See PDF for image]
Source: GAO.
[End of figure]
New Markets Tax Credit Program:
As illustrated in figure 5, under the New Markets Tax Credit program,
Treasury competitively awards tax credits to CDEs (such as a financial
institution or nonprofit organization), which offer the credits to
investors (including individuals, groups of investors, or
corporations). The equity or debt generated from such an offering is
used to finance eligible investments or projects, as described above.
In turn, these investments and projects may use New Markets Tax Credit
equity to leverage additional equity and debt to finance development or
related costs.
Figure 5: New Markets Tax Credit Basic Transaction Structure:
This figure is a flowchart showing new markets tax credit basic
transaction structure.
[See PDF for image]
Source: GAO.
[End of figure]
As illustrated in figure 6, rather than offering tax credits directly
to investors, under a leveraged transaction structure, a CDE may offer
credits to an investment fund. The investment fund pools equity
generated from the credit offering with other equity and debt, and
loans the funds to the CDE, which in turn makes qualified low-income
community investments.
Figure 6: New Markets Tax Credit Leveraged Transaction Structure:
This figure is a flowchart showing new markets tax credit leveraged
transaction structure.
[See PDF for image]
Source: GAO.
[End of figure]
[End of section]
Appendix III: Agency-reported Leverage Measures and Our Recalculations:
Table 4 outlines the leverage measures HUD and Treasury calculated for
the CDBG, HOME, HOPE VI, CDFI, Low-Income Housing Tax Credit, and New
Markets Tax Credit programs. Following table 4, we present several
leverage measure calculation scenarios for the selected programs.
Table 4: Agency-reported Leverage Measures by Program:
Program: HUD: CDBG;
Reported leverage measure[A, B]: 4.98:1[C].
Program: HUD: HOME;
Reported leverage measure[A, B]: 3.54:1.
Program: HUD: HOPE VI;
Reported leverage measure[A, B]: $634 million[D].
Program: Treasury: CDFI;
Reported leverage measure[A, B]: 26.82:1.
Program: Treasury: Low-Income Housing Tax Credits;
Reported leverage measure[A, B]: None[E].
Program: Treasury: New Markets Tax Credits;
Reported leverage measure[A, B]: 13.12:1[F].
Source: HUD and Treasury data.
[A] While HUD and Treasury generally reported leverage measures that
reflected the ratio of all other funds to program funds for each of the
selected programs, the data used to calculate them varied by program
and covered different periods because the programs have different
reporting schedules and requirements for funding recipients. The CDBG
ratio is based on data from Dec. 1, 2005, through May 1, 2007; the HOME
ratio is based on data from fiscal year 1992 through Aug. 31, 2007; the
HOPE VI measure is based on data from July 1, 2005, to Mar. 31, 2006;
the CDFI ratio is based on fiscal year 2005 data; and the New Markets
Tax Credit ratio is based on data from calendar years 2001 through
2005.
[B] All reported measures represent total program leverage. Treasury
also reports institutional and project measures for the CDFI and New
Markets Tax Credit programs--5.50 to 1 and 21.31 to 1, respectively,
for the CDFI program and 4.00 to 1 and 3.28 to 1, respectively, for the
New Markets Tax Credit program.
[C] HUD currently does not publicly report the leverage ratio it
reported to us for purposes of this report.
[D] Data on the amount of HOPE VI funding PHAs expended between July 1,
2005, and Mar. 31, 2006, were not available; therefore, we could not
convert the $634 million measure HUD reported into a leverage ratio.
[E] No agency collects leveraging data that could be used to calculate
a leverage measure for the Low-Income Housing Tax Credit program. As a
result, we contacted the housing finance agencies--the agencies that
are responsible for administering the program--to determine if they
collected leveraging data and whether they maintained their records
electronically. Of the 45 housing finance agencies that responded (79
percent response rate), 25 collect the dollar amounts contributed by
specific funding sources and keep that data electronically. We did not
review available data to determine its reliability or calculate an
estimated leverage measure for the program.
[F] Treasury currently does not publicly report the leverage measure it
reported to us for purposes of this report. The measure Treasury
reported for the New Markets Tax Credit program is not a leverage
ratio, but rather a money multiplier or multiplier ratio. A multiplier
ratio measures the total amount of investment $1 in tax credits
potentially can generate in low-income communities, whereas a leverage
ratio measures the additional amount of investment relative to a source
of funds (in this case, program funds, or the cost of the credit).
[End of table]
Recalculations of Agency-reported Leverage Measures:
Figures 7, 8, and 9 illustrate the leverage ratios HUD reported for the
CDBG, HOME, and HOPE VI programs, respectively (scenario A), and our
recalculations of the those measures to convey the ratio of nonfederal
funds to federal funds (scenario B) and public funds to private funds
(scenario C).[Footnote 72]
Based on cumulative data from December 1, 2005, to May 1, 2007, HUD
estimated the leverage ratio for the CDBG program to be 4.98 to 1
(other funds to CDBG program funds). However, we could not determine in
which category some CDBG funding data belonged; thus, we excluded these
data and revised the 4.98 to 1 ratio to 4.04 to 1 (scenario A in fig.
7). Using the revised dataset, we then recalculated the measure to
convey the ratio of nonfederal funds to federal funds (1.84 to 1,
scenario B in fig. 7) and the ratio of private funds to public funds
(0.55 to 1, scenario C in fig. 7).
Figure 7: Recalculated Leverage Measures for the CDBG Program:
This figure is a bar chart showing recalculated leverage measures for
the CDBG program.
[See PDF for image]
Source: GAO analysis of HUD calculations.
Notes: The agency-reported leverage ratio (4.98 to 1) shown in table 4
includes funds recorded in an "other funds" category, which represents
approximately 18 percent of all non-CDBG funds. According to HUD
officials, CDBG grantees use the other funds category for all financing
sources that do not have a specific data field, including Low-Income
Tax Credit equity. We excluded this category from our analysis in this
figure because it was not clear in which category(ies)--other federal,
state and local, or private--these funds actually belonged. Thus, for
purposes of this illustration, the leverage measure for the program is
4.04 to 1. The leverage ratios presented in this figure are for
illustrative purposes and do not represent actual leveraging values.
[End of figure]
Based on data on HOME activities completed in fiscal year 2006, HUD
estimated the leverage ratio for the HOME program to be 4.00 to 1 (all
other funds to HOME program funds). Using these same data, we
recalculated the measure to convey the ratio of nonfederal funds to
federal funds (1.15 to 1, scenario B in fig. 8), and the ratio of
private funds to public funds (0.62 to 1, scenario C in fig.
8).[Footnote 73]
Figure 8: Recalculated Leverage Measures for the HOME Program:
This figure is a bar chart showing recalculated leverage measures for
the HOME program.
[See PDF for image]
Source: GAO analysis of HUD calculations.
Note: For purposes of this analysis, we included Low-Income Housing Tax
Credit equity in the other federal funds category because the equity
represents forgone federal income and, therefore, is a direct cost to
the federal government. However, states and localities report Low-
Income Housing Tax Credit data and other federal funds data in separate
fields in HUD's database. The leverage ratios presented in this figure
are for illustrative purposes and do not represent actual leveraging
values.
[End of figure]
Using data on the 55 HOPE VI projects completed (that is, projects in
which all phases of construction were fully completed) as of March 31,
2006, we determined the leverage ratio of other funds to program funds
to be 1.13 to 1 (scenario A in fig. 9).[Footnote 74] Using the same
data, we recalculated the measure to convey the ratio of nonfederal
funds to federal funds (0.67 to 1, scenario B in fig. 9).
Figure 9: Recalculated Leverage Measures for the HOPE VI Program:
This figure is a bar chart showing recalculated leverage measures for
the HOPE VI program.
[See PDF for image]
Source: GAO analysis of HUF data.
Notes: We were unable to calculate a measure to express the ratio of
public to private funds in the HOPE VI program because the HOPE VI
database does not distinguish between private funds and other
nonfederal funds.
[End of figure]
In scenario B, Low-Income Housing Tax Credit equity is included in the
nonfederal funds category because the HOPE VI database tracks tax
credit equity in the nonfederal funds category. We generally consider
such equity as federal funds because it represents forgone federal
income and, therefore, a direct cost to the federal government--thus,
the extent of the federal investment in the program, as illustrated in
scenario B, is likely underestimated. We previously have found that Low-
Income Housing Tax Credit equity can account for a substantial
percentage of leveraged funds in a HOPE VI development. See GAO, Public
Housing: HOPE VI Leveraging Has Increased, but HUD Has Not Met Annual
Reporting Requirement, GAO-03-91 (Washington, D.C.: Nov. 15, 2002).
Unlike the HUD programs, we generally were not able to recalculate the
leverage measures Treasury reported for the CDFI and New Markets Tax
credit programs to convey the extent of public and private investment
in them because available leveraging data did not distinguish between
or correctly categorize public and private contributions in either
program. However, we were able to deconstruct the measures to
approximate the extent of institutional and project leverage in them
(see figs. 10 and 11).[Footnote 75]
In the CDFI program, match leverage (that is, the ratio of nonfederal
match funds to program funds) represented approximately one-fifth of
total institutional leverage, while debt leverage (that is, additional
private debt CDFIs were able to attract with program funds and matched
funds combined) represented four-fifths of total institutional
leverage, according to Treasury data.[Footnote 76] A majority of the
total leverage that occurred in the CDFI program in fiscal year 2005
occurred at the project level--according to Treasury data, for every $1
of federal funds the agency contributed to CDFIs, they were able to
leverage an additional $21.31 at the project level (see fig.
10).[Footnote 77]
Figure 10: Institutional and Project Leverage in the CDFI Program:
This figure is a bar chart showing institutional and project leverage
in the CDFI program.
[See PDF for image]
Source: GAO analysis of Treasury data.
Notes: Treasury assumed match leverage to be 1 to 1, even though CDFIs
may attract more than $1 in nonfederal funds for every $1 received in
program funds. Because match data were not complete, this calculation
only approximates the extent of leveraging in the program.
Match leverage is the ratio of match funds to program funds. Debt
leverage is equal to liabilities/net assets multiplied by match
leverage. Project leverage is equal to (match leverage + debt leverage)
x (total project costs/total loan origination amounts).
[End of figure]
In the New Markets Tax Credit program, institutional leverage (that is,
net equity generated through the offering of the credits) represented
approximately 58 percent of total leveraging that occurred in the
program, while project leverage represented approximately 42 percent of
total leveraging (see fig.11).[Footnote 78]
Figure 11: Institutional and Project Leverage in the New Markets Tax
Credit Program:
This figure is a bar chart showing institutional and project leverage
in the new markets tax credit program.
[See PDF for image]
Source: GAO.
Note: Of the 538 projects for which Treasury had project-level
leveraging data, 399, or 74 percent reported data in the public sources
field (either $0 or a positive amount); 139, or 26 percent left the
field blank. Because the public sources of data were not complete, this
calculation only approximates the extent of public investment in the
program.
The leverage measure(s) reported in the figure differs from Treasury's
reported measure in table 1--the 13.12 to 1 ratio Treasury computed for
the New Markets Tax Credit program overstates the extent of leveraging
that occurred in the program because it describes the total amount of
investment $1 in tax credits potentially can generate in low-income
communities, rather than the additional amount of investment directly
acquired as a result of receiving other sources of funding (for
example, program funds, or in the case of the New Markets Tax Credit
program, the cost of the credit or forgone federal revenue). Using
Treasury data, we computed a leverage measure for the program, which is
presented in the figure.
[End of figure]
[End of section]
Appendix IV: Housing and Community and Economic Development Project
Profiles:
To determine how federal funds have been leveraged in HUD's CDBG, HOME,
and HOPE VI programs and Treasury's CDFI, Low-Income Housing Tax
Credit, and New Markets Tax Credit programs, we toured and obtained
information on a number of projects that used funds from the selected
programs in combination with other federal, state, local, and private
funds for housing and community and economic development. Appendix I
describes how we selected the communities to visit and the projects to
include in this report.
HUD's CDBG, HOME, and HOPE VI programs and Treasury's CDFI, Low-Income
Housing Tax Credit, and New Markets Tax Credit programs are among a
number of federal programs that fund housing and community and economic
development.[Footnote 79] Specific information about the features of
these programs follows:
* CDBG provides annual grants on a formula basis to entitlement
communities and states to develop viable urban communities by providing
decent housing and a suitable living environment, and by expanding
opportunities, principally for low-and moderate-income persons. Under
CDBG, communities and states develop their own programs and funding
priorities.
* HOME provides formula grants to states and localities to fund a range
of activities that buy, build, and rehabilitate affordable housing for
rent or homeownership or provide direct rental assistance to low-income
households.
* HOPE VI is part of HUD's effort to transform public housing by
providing grants that fund the demolition of severely distressed public
housing; the capital costs of major rehabilitation, new construction,
and other physical improvements; and other resident-related services.
* Through the CDFI program, Treasury provides CDFIs with financial
assistance in the form of grants, loans, equity investments, and
deposits to enhance their ability to make loans and investments and
provide services for the benefit of low-income communities and persons.
CDFI funds can be used for economic development, affordable housing,
and community development financial services.
* Under the Low-Income Housing Tax Credit program, states are
authorized to allocate federal tax credits as an incentive to private
investors to develop rental housing for low-income households. The
equity generated by the sale of tax credits is used to lower the
financing costs of housing developments by reducing the debt or equity
the developer otherwise would incur or contribute.
* The New Markets Tax Credit program permits taxpayers to receive a
credit against federal income taxes for making qualified equity
investments in CDEs. CDEs use the equity generated by the sale of the
credits to make investments in qualified low-income businesses.
Figures 12 through 14 illustrate several projects that leveraged
federal funds for the development of affordable housing. Figures 15
through 20 illustrate several projects that leveraged federal funds for
community and economic development activities.
Figure 12: Hilltop Oaks Apartments, San Antonio, Texas:
This figure is a photograph of Hilltop Oaks Apartments in San Antonio,
Texas with text.
[See PDF for image]
Sources: SAAHC (information); GAO (analysis, photo).
[A] All units in the development are reserved for individuals and
families who earn 60 percent or less of the area median income.
[B] SAAHC used 4 percent Low-Income Housing Tax Credits to help finance
the Hilltop Oaks Apartments project.
[C] The City of San Antonio Rental Rehabilitation loan primarily
consists of HOME funds.
[End of figure]
Figure 13: Near North Apartments, Chicago, Illinois:
This figure is a photograph with text of Near North Apartments in
Chicago, Illinois.
[See PDF for image]
Sources: City of Chicago Department of Housing (information); GAO
analysis, photo).
[A] The Shelter Plus Care program provides rental assistance for hard-
to-serve homeless persons with disabilities in conjunction with
supportive services funded from sources outside the program. Project-
based Section 8 vouchers are a component of a public housing agency's
Housing Choice Voucher program. A housing agency can attach up to 20
percent of its voucher assistance to specific housing units if the
owner agrees to either rehabilitate or construct the units, or set
aside a portion of the units in an existing development.
[B] The Leadership in Energy and Environmental Design (LEED) Green
Building Rating System is the nationally accepted benchmark for the
design, construction, and operation of high-performance green
buildings. To earn certification, a building project must meet certain
prerequisites and performance benchmarks (credits) within each
category. Projects are awarded Certified, Silver, Gold, or Platinum
certification depending on the number of credits they achieve.
[C] This amount represents the total development cost minus the
donations made by the city of Chicago and suppliers.
[D] This amount represents the value of the donation. The city of
Chicago sold the land to the developer for $1.
[E] This amount represents the value of the donation.
[F] The total includes the value of all donations. Total development
costs less donations were equal to $13,664,000.
[G] Percentages do not add to 100 due to rounding.
[End of figure]
Figure 14: New Columbia, Portland, Oregon:
This figure is a photograph of three apartment complexes in New
Columbia, Portland Oregon.
[See PDF for image]
Sources: The Housing Authority of Portland and Enterprise Community
Partners Inc (information); GAO (analysis, photos).
[A] The Leadership in Energy and Environmental Design (LEED) Green
Building Rating System is the nationally accepted benchmark for the
design, construction, and operation of high-performance green
buildings. To earn certification, a building project must meet certain
prerequisites and performance benchmarks (credits) within each
category. Projects are awarded Certified, Silver, Gold, or Platinum
certification depending on the number of credits they achieve.
[B] Percentages do not add to 100 due to rounding.
[C] Enterprise Community Partners, Inc. formed a partnership with Bank
of America to facilitate the allocation of the New Markets Tax Credits.
[End of figure]
Figure 15: ACCION Texas, San Antonio, Texas:
This figure is a combination of photographs with text from ACCION
Texas, San Antonio, Texas.
[See PDF for image]
Source: ACCION Texas (information); GAO (analysis, photos).
[A] The average credit score for ACCION Texas's clients is 570, which
generally is below what is required to obtain credit in the commercial
market.
[B] A colonia is an unincorporated community along the U.S.-Mexico
border.
[End of figure]
Figure 16: Avenue North, Philadelphia, Pennsylvania:
This figure is a combination of photographs with text showing Avenue
North, Philadelphia, Pennsylvania.
[See PDF for image]
Sources: The Reinvestment Fund and GAO (analysis, photographs).
[A] The Reinvestment Fund is a certified CDE and CDFI. The Fund
received New Markets Tax Credit allocations in 2003 and 2006 totaling
approximately $113.5 million.
[B] The Reinvestment Fund provided the New Markets Tax Credit
allocation, and Citizens Bank was the New Markets Tax Credit investor.
Citizens Bank agreed to assume The Reinvestment Fund's debt after the 7-
year tax credit term expired and also agreed to keep its loans below
the market rate for 15 years.
[End of figure]
Figure 17: Gerding Theater at the Armory, Portland, Oregon:
This figure is a picture with text of Gerding Theater at the Armory,
Portland, Oregon.
[See PDF for image]
Sources: The Reinvestment Fund and GAO (analysis, photographs).
[A] The Enterprise Community program (part of the larger Empowerment
Zone/Enterprise Community program) is a large-scale federal effort
directed at the revitalization of impoverished rural and urban
communities. The program provides grants to public and private entities
for social services and community redevelopment and tax benefits to
local businesses to attract or retain jobs and businesses in distressed
communities. The State of Oregon's Urban Renewal program is a state-
authorized redevelopment and finance program designed to help
communities improve and redevelop areas that are physically
deteriorated, suffering economic stagnation, unsafe, or poorly planned.
[B] The Leadership in Energy and Environmental Design (LEED) Green
Building Rating System is the nationally accepted benchmark for the
design, construction, and operation of high-performance green
buildings. To earn certification, a building project must meet certain
prerequisites and performance benchmarks (credits) within each
category. Projects are awarded Certified, Silver, Gold, or Platinum
certification depending on the number of credits they achieve.
[C] Percentages do not add to 100 due to rounding.
[End of figure]
Figure 18: New Hampshire Community Loan Fund Manufactured Housing
Program: Exeter River Cooperative, Exeter, New Hampshire:
This figure is a photograph with text of New Hampshire Community Loan
Fund Manufactured Housing Program, Exeter River Cooperative, Exeter,
New Hampshire.
[See PDF for image]
Source: The New Hampshire Community Loan Fund (information); GAO
(analysis, photo).
[A] As a certified CDFI, the New Hampshire Community Loan Fund has
received assistance from Treasury. As required under the CDFI program,
the Loan Fund matched (leveraged) this assistance with nonfederal
funds.
[B] N.H. Rev. Stat. sec. 205-A:21.
[C] Percentages do not add to 100 due to rounding.
[End of figure]
Figure 19: Seattle Cooperative Children's Center, Seattle, Washington:
This figure is a photograph with text of Seattle Cooperative Children's
Center in Seattle Washington.
[See PDF for image]
Source: Shorebank Enterprise Cascadia (information); GAO analysis,
photo).
[A] As a certified CDFI, Shorebank Enterprise Cascadia has received
assistance from Treasury. As required under the CDFI program, Shorebank
Enterprise Cascadia matched (leveraged) this assistance with nonfederal
funds.
[End of figure]
Figure 20: Shoalwater Bay Wellness Center, Tokeland, Washington:
This figure is a photograph with text of the Shoalwater Bay Wellness
Center in Tokeland Washington.
[See PDF for image]
Source: Shorebank Enterprise Cascadia (information); GAO (analysis,
photos).
[A] The Shoalwater Bay Tribe resides on the coast of Washington at
Willapa Bay, near Tokeland, Washington. The original Shoalwater
Reservation was established on September 22, 1866; the tribe gained
federal recognition on March 10, 1971, and adopted a constitution and
elected a tribal council shortly thereafter.
[End of figure]
[End of section]
Appendix V Comments from the Department of Housing and Urban
Development, Office of Community Planning and Development:
U.S. Department Of Housing And Urban Development:
Washington, DC 20410-7000:
Office Of Community Planning And Development:
December 6, 2007:
Mr. William B. Shear:
Director, Financial Markets and Community Investment:
U.S. Government Accountability Office 441 G Street, NW:
Washington, DC 21548:
Dear Mr. Shear:
This letter provides the comments of HUD's Office of Community Planning
and Development (CPD) on the draft GAO report entitled "More
Information on Leverage Measures' Accuracy and Linkage to Program Goals
Is Needed in Assessing Performance" (GAO-08-136). The Department
appreciates this opportunity to respond to the draft report.
In general, CPD is pleased with the results presented in the draft
report but does have several comments that should be incorporated in
order to highlight certain facts. CPD requests that a strong statement
be made early in the report that neither the Community Development
Block Grant (CDBG) nor HOME Investment Partnerships (HOME) programs
have statutory or regulatory provisions requiring leverage of other
funding sources. Ideally this statement would be included on page four,
paragraph two, where the CDBG and HOME programs are identified as
subjects of the draft report. We recognize that such statements are
made at various points in the draft report but desire to ensure that
this critical fact is conveyed as early as possible and as often as
possible.
As CPD does not have a formal leverage measure for the CDBG program, we
are concerned with the language on page six, paragraph one, sentence
one, that "[T]he leverage measures . lacked transparency because the
agencies generally did not disclose the limitations of the data or
methods used to calculate them." Further, the draft report contains, in
multiple places, similar language which leaves the impression that HUD
provides a CDBG leverage measure. CPD asks that all such instances be
revised in some manner to reflect that CPD does not report a CDBG
leverage measure.
CPD is also concerned with the same language regarding the HOME
program. HOME participating jurisdictions may voluntarily report
leveraging on a project-level basis to demonstrate the program's
ability to leverage other public and private funding. The results of
this reporting are found on the HOME Program National Production Report
and SNAPSHOT of HOME Program Performance. CPD identifies the source of
the data and limitations of the calculations used to determine the
amount of leverage on the National Production Report and the SNAPSHOT
report's accompanying "Explanation of Performance Categories."
Page 7 of the report contends that some funding recipients use
leveraging to prioritize funding decisions "giving funding priority to
projects that leverage more over those that leverage less, but which
may fill a greater or more immediate need within a community". CPD
formula programs emphasize local decision making, meaning CPD does not
approve CDBG grantees' or HOME participating jurisdictions' project
funding decisions. Moreover, CDBG has no leveraging measure on which to
base local funding decisions and while HOME does report leveraging
ratios for individual participating jurisdictions, the ratio is based
solely on completed rental projects, which is one of the four eligible
activity types of HOME. Therefore, CPD requests that the language on
page 7 and all other instances where it is implied or stated that
leveraging affects funding decisions for CDBG and HOME be modified to
indicate that funding decisions are made locally and that leveraged
funds on a project level basis are not approved by CPD.
Footnote 32 on page 18 of the draft report indicates that HUD will
publish a CDBG leverage measure by the end of calendar year 2007. This
statement is repeated in footnote 54 on page 28. It was CPD's intent to
convey that, beginning with program year 2006 CDBG grantees profiles,
CPD would post the leverage amount reported by each grantee. These
grantee profiles, posted on HUD's website, provide summary information
with respect to grantee uses of CDBG funds during the program year.
More than 80 percent of program year 2006 profiles for CDBG grantees
will be posted as of December 31, 2007 and will reflect funds leveraged
as reported by the grantees. In some cases, the amount leveraged may be
zero as grantees either failed to report or did not leverage other
funding sources.
Further, it is CPD's intent to continue to work during FY 2008 to
increase the percentage of CDBG grantees that voluntarily report
leverage data and other program information through IDIS. Concurrently,
CPD will attempt to improve the quality of leverage data through closer
analysis of the data provided by grantees. CPD proposes to follow upon
this GAO report to produce a program-wide leverage measure for the CDBG
program in the fall of 2008. This effort will build upon improvements
in the level and quality of grantee reporting during FY 2008. This
effort would, in part, address the Recommendations for Executive Action
presented on page 36 of the draft report.
Footnote 54 on page 28 addresses HUD's view on the value of leveraging
measures to assess program performance. CPD recommends that the opening
clause of the first sentence be deleted because it is not relevant to
the point being made.
CPD also finds GAO's recalculation of HOME leveraging data in Appendix
III, purportedly based on 2006 data provided by the Office of
Affordable Housing Programs, unclear. Different results may be obtained
depending on the specific criteria applied to the data in performing
the 3
recalculation. Specifically, it is not evident from the report if data
on HOME projects funded in 2006, completed in 2006, or funded with 2006
HOME dollars was used in the recalculation. CPD asks that GAO clarify
the data criteria used in the recalculation.
If you have any questions regarding this response, please contact me at
(202) 708-3587.
Sincerely,
Signed by:
Richard J. Kennedy:
Acting Deputy Assistant Secretary for Grant Programs:
[End of section]
Appendix VI: Comments from the Department of the Treasury, Community
Development Financial Institutions Fund:
Department Of The Treasury:
Community Development Financial Institutions Fund:
601 Thirteenth Street, NW, Suite 200 South:
Washington, DC 20005:
December 11, 2007:
Mr. William B. Shear:
Director, Financial Markets and Community Investments:
U.S. Government Accountability Office:
441 G Street, N.W.:
Washington, D.C. 20548:
Dear Mr. Shear:
Thank you for providing the Community Development Financial
Institutions (CDFI) Fund, an office within the U.S. Department of
Treasury, with the opportunity to comment on the draft Government
Accountability Office (GAO) report, "HUD and Treasury Programs: More
Information on Leverage Measures' Accuracy and Linkage to Program Goals
Is Needed in Assessing Performance."
Before addressing specific elements of the draft report, I want to
acknowledge the professionalism of the GAO staff; particularly the
efforts that they have made to understand our programs. We appreciate
the GAO's finding that each of the three Treasury programs (the Low-
Income Housing Tax Credit Program; the New Markets Tax Credit Program;
the CDFI Program) was "designed to leverage other funds in a number of
ways and, as a result, leveraging directly relates to each program's
goals and core activities and leverage measures can be used to describe
program outputs." We share your belief that this leveraging component
sets these programs apart from other programs that you may have
evaluated and, indeed, from many other Federal government programs.
With respect to other elements and findings of your report, we do have
several comments to offer:
1. Disclosure of methodology for determining leverage figures. The GAO
draft report states that, when presenting leverage measures for the
selected programs in performance and budget reports, and in other
resources, "The leverage measures HUD and Treasury reported for the
selected programs lacked transparency because the agencies generally
did not disclose the limitations of the data or the methods used to
calculate them" (page 6). The CDFI Fund cannot comment on the data
collection or disclosure practice of the other agencies participating
in this study. However, with respect to the CDFI Program, the CDFI Fund
has publicly disclosed its methodologies on multiple occasions --
including reports issued as early as March of 2005 ("CDFIs Leverage
CDFI Program Awards Nearly $20 to $1"), and as recently as July of 2007
("Growth, Diversity, Impact: A Snapshot of CDFIs in FY 2003").
With respect to the NMTC Program, the CDFI Fund has not provided
information relating to the leveraging of program dollars in
performance and budget reports or in any other public sources. It is
therefore inappropriate to include the NMTC Program as part of this
report's findings relating to the public disclosure of data limitations
or methodologies � since NMTC leverage data has never before been
publicly released by the CDFI Fund. That being said, to the extent the
CDFI Fund does release leveraging figures in the future, it will make
every effort to include a discussion of data methodologies and
limitations.
2. Incomplete Reporting of Data. The draft GAO report states that "the
leverage measures the agencies reported for the selected programs were
based on incomplete data and thus did not capture the actual extent of
leveraging in the programs." (pg. 15). With respect to the CDFI
Program, the draft report notes that there may actually be additional
leveraging at the institutional level that occurs beyond the 1:1 match
of CDFI Program award dollars, since many awardees procure nonfederal
funds that exceed their final award amounts. The statutory requirement
of the CDFI Program is that awardees produce 1:1 ratio of nonfederal
matching funds for CDFI Program awards. Therefore, the CDFI Fund
collects the matching fund data necessary to ensure that this statutory
requirement is satisfied. The CDFI Fund is aware that the actual
leveraging of CDFI Program awards may exceed the 1:1 ratio, but does
not feel it is appropriate to include excess matching fund sources as
part of the CDFI Fund's leveraging calculations.
The draft GAO report includes two observations with respect to the CDFI
Fund's reporting of NMTC leveraging data:
a. Project-level data was unavailable for 26 percent of NMTC projects,
which could lead to a potential underestimation of total leveraging
(page 16). The CDFI Fund collects project-level data on every NMTC
transaction. NMTC Program participants are asked to indicate the total
project costs, including both NMTC contributions and other
contributions. Through the initial four years of data collection under
the NMTC Program, this project cost data field was optional, not
mandatory (the requirements for data submissions have since been
changed to make this a mandatory field). The draft GAO report indicates
that the missing project data values from the earlier years could
result in underestimating total leverage across all of the years of the
program. However, the CDFI Fund believes that the 74 percent of
projects for which data was available is sufficiently representative of
the entire pool of NMTC transactions to support its determination that,
on average, each $1 of NMTC investment proceeds supports a total
investment of $3.28 at the project level.
b. Treasury assumed that CDEs contribute 100 percent of tax credit
equity to qualified low-income community investments, even though CDEs
are permitted to retain up to 15 percent for administrative and other
purposes (page 16). This is a valid observation. However, the CDFI Fund
does not believe that its:
c. leveraging calculations would be substantially different if the non-
invested proceeds were excluded. In order to increase the likelihood of
receiving an allocation award, most CDEs commit to investing
significantly more than the minimally required 85 percent of tax credit
proceeds into projects. In the 2007 allocation round, for example, 60
of the 61 awardees indicated that at least 95 percent of their tax
credit proceeds would be invested in qualified projects, including 49
that committed to investing at least 97 percent of their proceeds into
qualifying projects.
3. The NMTC leveraging ratio of 13.12:1 is not a "leverage" ratio, but
rather a "multiplier" ratio. The CDFI Fund believes that its
presentation of the NMTC leverage is appropriate and consistent with
GAO guidance. As stated in the draft GAO report, leveraging can be
defined as "using a relatively small amount of federal funds to attract
private investment." (page 3). In the context of a tax credit program,
where federal funds are not directly provided to a project, the federal
funds are represented by foregone tax revenue. In the case of the NMTC
Program, it has been estimated that each $1 of equity invested in a
project results in a net tax revenue loss to the Federal government of
25 cents [Footnote 80]. As stated above, this 25 cent investment, on
average, results in a total of $3.28 invested at the project level �
for a true leveraging ratio of over 13:1 [$3.28/.25 = 13.12].
We thank you for the opportunity to share these comments with you, and
hope that they will inform your preparation of the final report.
Sincerely,
Signed by:
Donna Gambrell:
Director:
[End of section]
Appendix VII: GAO Contact and Staff Acknowledgments:
GAO Contact:
William B. Shear, (202) 512-8678 or [email protected]:
Acknowledgments:
In addition to the contact named above Marianne Anderson, William
Bates, Elizabeth Curda, Daniel Garcia-Diaz, Terence Lam, Alison Martin,
John McGrail, Jackie Nowicki, Josephine Perez, Barbara Roesmann, Cory
Roman, and Charles Wilson Jr. made key contributions to this report.
[End of section]
Related GAO Products:
Leveraging Federal Funds for Housing, Community, and Economic
Development. GAO-07-768R. Washington, D.C.: May 25, 2007.
Performance Budgeting: PART Focuses Attention on Program Performance,
but More Can Be Done to Engage Congress. GAO-06-28. Washington, D.C.:
October 28, 2005.
Performance Budgeting: Observations of the Use of OMB's Program
Assessment Rating Tool for the Fiscal Year 2004 Budget. GAO-04-174.
Washington, D.C.: January 30, 2004.
Tax Administration: IRS Needs to Further Refine Its Tax Filing Season
Performance Measures. GAO-03-143. Washington, D.C.: November 22, 2002.
Public Housing: HOPE VI Leveraging Has Increased, but HUD Has Not Met
Annual Reporting Requirement. GAO-03-91. Washington, D.C.: November 15,
2002.
Performance Reporting: Few Agencies Reported on the Completeness and
Reliability of Performance Data. GAO-02-372. Washington, D.C.: April
26, 2002.
Managing for Results: Challenges Agencies Face in Producing Credible
Performance Information. GAO/GGD-00-52. Washington, D.C.: February 4,
2000.
Performance Plans: Selected Approaches for Verification and Validation
of Agency Performance Information. GAO/GGD-99-139. Washington, D.C.:
July 30, 1999.
Agency Performance Plans: Examples of Practices that Can Improve
Usefulness to Decisionmakers. GAO/GGD/AIMD-99-69. Washington, D.C.:
February 26, 1999.
Grant Programs: Design Features Shape Flexibility, Accountability, and
Performance Information. GAO/GGD-98-137. Washington, D.C.: June 22,
1998.
Managing for Results: Analytic Challenges in Measuring Performance.
GAO/HEHS/GGD-97-138. Washington, D.C.: May 30, 1997.
[End of section]
Footnotes:
[1] Leverage measures--for example, leverage ratios, total dollars
leveraged, and leverage factors--provide information on the extent to
which a program or project has been successful in attracting or
combining other funds.
[2] Pub. L. No. 103-62, 31 U.S.C. 1115 et seq.
[3] As mentioned previously, it is in these types of reports that
agencies often report leverage measures for their programs.
[4] See GAO, Managing for Results: Analytic Challenges in Measuring
Performance, GAO/HEHS/GGD-97-138 (Washington, D.C.: May 30, 1997).
[5] See Executive Office of the President, Office of Management and
Budget, The President's Management Agenda (Washington, D.C., 2002).
[6] See GAO, Performance Budgeting: PART Focuses Attention on Program
Performance, but More Can Be Done to Engage Congress, GAO-06-28
(Washington, D.C.: Oct. 28, 2005).
[7] See GAO, Leveraging Federal Funds for Housing, Community, and
Economic Development, GAO-07-768R (Washington, D.C.: May 25, 2007).
[8] Our first report, GAO-07-768R, discussed stakeholder perspectives
on the use, measurement, and implications of leveraging federal funds
for housing and community and economic development as well as the types
of data the Department of Housing and Urban Development collects that
could be used to measure the extent of leveraging in its Section 108
Loan Guarantee program.
[9] The background section of this report provides information on the
purpose and scope of these programs. Throughout this report, we refer
to the CDFI Financial Assistance program as the CDFI program. As
discussed in the background section of this report, the CDBG and HOME
programs do not have statutory or regulatory leveraging requirements.
[10] See GAO, Assessing the Reliability of Computer-Processed Data, GAO-
02-15G (Washington, D.C.: Sept. 2002). As explained in further detail
in app. I, we were unable to determine the reliability of HUD's CDBG
and HOME leveraging data.
[11] A budget justification is the set of documents an agency submits
to congressional appropriation committees in support of its budget
request.
[12] We and Treasury both noted that the agency currently does not
publish a measure for the New Markets Tax Credit program.
[13] See GAO-07-768R.
[14] The CDBG program was authorized under Title I of the Housing and
Community Development Act of 1974 (Pub. L. 93-383, as amended, 42
U.S.C. 5301 et seq.) In 1981, Congress amended the act to allow states
the opportunity to administer CDBG funds for nonentitlement
communities--units of general local government that do not receive CDBG
funds directly from HUD as part of the entitlement program.
[15] The State of Hawaii permanently has elected not to receive CDBG
program funding. As a result, HUD awards these state-level funds
directly to Hawaii's three nonentitlement communities.
[16] HOME was authorized under Title II of the Cranston-Gonzalez
National Affordable Housing Act of 1990, as amended (Pub. L. 101-625,
42 U.S.C. 12721, et seq.) States administer HOME funds for localities
that do not qualify to receive allocations directly from HUD.
[17] Under a tenant-based rental assistance program, states and
localities allocate HOME funds to eligible households for the payment
of rent on units of their choosing.
[18] HOME matching contribution requirements are outlined in 24 C.F.R.
92.218-92.222. See also, U.S. Department of Housing and Urban
Development Community Development and Planning, HOME Program Match
Guidance, Notice: CPD 97-03 (Washington, D.C., Mar. 27, 1997).
[19] In 1989, Congress created the National Commission on Severely
Distressed Public Housing (Pub. L. 101-235) to assess the state of the
nation's public housing and make recommendations for its improvement.
The commission's 1992 report to Congress--see National Commission of
Severely Distressed Public Housing, Final Report of the National
Commission on Severely Distressed Public Housing (Washington, D.C.,
Aug. 1992)--recommended the establishment of a demonstration program to
implement its proposals for change, which included addressing the needs
of public housing residents and improving the physical conditions of
public housing. Congress authorized and funded the Urban Revitalization
Demonstration Program, or HOPE VI, through annual appropriations bills
until 1998, when HOPE VI was authorized through fiscal year 2002 under
� 535 of the Quality Housing and Work Responsibility Act of 1998, as
amended (42 U.S.C. 1437v), and subsequently reauthorized through fiscal
year 2008.
[20] HUD outlines HOPE VI match and leverage requirements in annual
notices of funding availability. For example, see "Supplement to the
Fiscal Year (FY) 2006 SuperNOFA for HUD's Discretionary Programs: NOFAs
for the HOPE VI Revitalization Grants Program and HOPE VI Main Street
Grants Program; Notice," 71 Federal Register 18496-18560 (Apr. 11,
2006). Leverage is one of several rating factors HUD considers in the
HOPE VI application process. Other rating factors include capacity of
the development team, need, resident and community involvement,
community and supportive services, relocation, fair housing and equal
opportunity, well-functioning communities, and soundness of approach.
[21] Applications for fiscal year 2007 HOPE VI funding were due Nov. 7,
2007.
[22] The CDFI Fund within Treasury administers the CDFI program. The
Fund was established as a bipartisan initiative under the Riegle
Community Development and Regulatory Improvement Act of 1994 (Pub. L.
103-325, 12 U.S.C. 4701 et seq.) to promote economic revitalization and
community development through investment in and assistance to CDFIs
through several programs, including the CDFI and Bank Enterprise Award
programs. Investment areas and targeted populations are defined in 12
C.F.R. � 1805.201(b)(3). CDFIs are private profit-making and nonprofit
financial institutions that provide financial services to distressed
geographic areas and populations that are underserved by conventional
lenders and investors. Common types of CDFI organizations include
community development banks, community development credit unions,
nonprofit community development loan funds, microenterprise loan funds,
and community development venture capital funds.
[23] CDFIs also may use their program and match funds to leverage
additional debt from banks and other lending institutions. Together,
program funds, match funds, and debt comprise institutional leverage in
the program. At the project level, funding recipients may use grants or
loans from CDFIs to leverage funds from other public and private
sources to finance project costs. App. II describes in more detail how
leveraging occurs in the CDFI program.
[24] Before receiving any financial assistance through the CDFI
program, a CDFI must be certified by Treasury; that is, meet six
statutory and regulatory criteria: (1) have a primary mission of
promoting community development; (2) principally serve an investment
area or targeted population; (3) be an insured depository institution,
or make loans or development investments as its predominant business
activity; (4) provide development services--such as technical
assistance or counseling--with its financing activity; (5) maintain
accountability to its target market; and (6) be a nongovernmental
entity and not be controlled by any governmental entities.
[25] The Tax Reform Act of 1986 (Pub. L. No. 99-514, 26 U.S.C. 42)
authorized the Low-Income Housing Tax Credit program.
[26] App. II describes in more detail how leveraging occurs in the Low-
Income Housing Tax Credit program.
[27] See 26 U.S.C. 42(h)(3)(C) and Rev. Proc. 2006-53, 2006-48 I.R.B.
996. Low-Income Housing Tax Credits are offered at two rates, 9 percent
and 4 percent. Most new construction and substantial rehabilitation
projects are eligible for 9 percent credits. Projects that are financed
through the issuance of tax-exempt bonds automatically may qualify for
4 percent credits. Credits awarded to these projects are not subject to
the per capita limit; however, the underlying bonds are subject to the
state private activity bond cap.
[28] The Community Renewal Tax Relief Act of 2000 (Pub. L. No. 106-554,
26 U.S.C. 45D) authorized the New Markets Tax Credit program, which
Treasury's CDFI Fund administers. A CDE is a domestic corporation or
partnership that is an intermediary vehicle for the provision of loans,
investments, or financial counseling in low-income communities through
the New Markets Tax Credit program. CDEs must be certified as such by
Treasury. App. II describes in more detail how leveraging occurs in the
New Markets Tax Credit program.
[29] See "Notice of Allocation Availability Inviting Applicants for the
CY 2007 Allocation Round of the New Markets Tax Credit Program," 71
Federal Register, 70835, 70841 (Dec. 6, 2006).
[30] For additional information on the effect of data limitations on
performance measurement, see GAO, Performance Reporting: Few Agencies
Reported on the Completeness and Reliability of Performance Data, GAO-
02-372 (Washington, D.C.: Apr. 26, 2002); Managing for Results:
Challenges Agencies Face in Producing Credible Performance Information,
GAO/GGD-00-52 (Washington, D.C.: Feb. 4, 2000); Performance Plans:
Selected Approaches for Verification and Validation of Agency
Performance Information, GAO/GGD-99-139 (Washington, D.C.: July 30,
1999); and Agency Performance Plans: Examples of Practices That Can
Improve Usefulness to Decisionmakers, GAO/GGD/AIMD-99-69 (Washington,
D.C.: Feb. 26, 1999).
[31] App. III describes the leverage measures HUD and Treasury reported
for each of the selected programs. HUD and Treasury currently do not
publish leverage measures for the CDBG and New Markets Tax Credit
programs, respectively; however, officials from both agencies said that
they plan to publish measures for the programs in the near future. In
comments on this report, HUD clarified that the agency will post
program year 2006 leveraging and other CDBG data on its Web site in
grantee profiles by Dec. 31, 2007. According to HUD, these profiles
will reflect funds leveraged as reported by CDBG grantees. In some
cases, the amount leveraged may be zero as grantees either failed to
report or did not leverage other funding sources.
[32] See The Community Development Financial Institutions Fund, U.S.
Department of the Treasury, "CDFIs Leverage CDFI Program Awards Nearly
$27 to $1" (Washington, D.C., Feb. 13, 2007) and The Community
Development Financial Institutions Fund, U.S. Department of the
Treasury, "CDFIs Leverage CDFI Program Awards Nearly $20 to $1!"
(Washington, D.C., Mar. 2005). In comments on this report, Treasury
noted that the agency also disclosed data limitations in Community
Development Financial Institutions Fund, "Growth, Diversity, Impact: A
Snapshot of CDFIs in FY 2003" (Washington, D.C., June 1, 2007).
[33] HUD and Treasury currently do not publish leverage measures for
the CDBG and New Markets Tax Credit programs, respectively; however,
officials from both agencies said that they plan to publish measures
for the programs in the near future.
[34] For example, see GAO-02-372, GAO/GGD-00-52, GAO/GGD-99-139, and
GAO/GGD/AIMD-99-69.
[35] See GAO/GGD-99-139.
[36] For example, see Office of Management and Budget, Performance
Measurement Challenges and Strategies (Washington, D.C., June 18,
2003). Federal agencies also are expected to adhere to OMB guidelines
for ensuring and maximizing the quality of the data they report to the
public. See "Guidelines for Ensuring and Maximizing the Quality,
Objectivity, Utility, and Integrity of Information Disseminated by
Federal Agencies; Notice; Republication," 67 Federal Register 8452-8460
(Feb. 22, 2002).
[37] According to agency officials, Treasury plans to report a measure
for the New Markets Tax Credit program that describes the total
potential amount of investment generated by tax credits. However,
Treasury's approach for calculating this measure is different from the
approach used by other programs when measuring leveraging. App. III
discusses the differences in Treasury's approach.
[38] App. III describes the leverage measures HUD and Treasury reported
for each of the selected programs.
[39] Investment risk is the potential for fluctuation in the value of
an investment, which could result in loss of principal.
[40] The background section and app. II provide information on how
these programs leverage.
[41] See GAO-07-768R. At the institutional level, a CDFI, for example,
groups together projects with varying levels of risk in a diversified
portfolio, which hedges risks associated with any particular project or
type of project.
[42] HUD's and Treasury's ability to recalculate leverage measures in
the manner described in fig. 1 depends on how they collect leveraging
data and on their method of calculation. As described in app. III, we
were not able to recalculate the leverage measures of some of the
selected programs to reflect all of the scenarios in fig. 1.
[43] For example, see GAO-02-372, GAO/GGD-00-52, GAO/GGD-99-139, and
GAO/GGD/AIMD-99-69, and Office of Management and Budget, Performance
Challenges and Strategies (Washington, D.C., June 18, 2003).
[44] See GAO, Tax Administration: IRS Needs to Further Refine Its Tax
Filing Season Performance Measures, GAO-03-143 (Washington, D.C.: Nov.
22, 2002). We also noted that measures should be clearly stated,
consistent with the methodologies used to calculate them, and balanced.
(Typically, agencies develop a suite of goals and measures covering the
various priorities of their programs. Balance exists when the suite of
measures covers those priorities.)
[45] For measurement purposes, all or a portion of tax credit equity is
considered a federal source of funds, as (all or a portion of) the
equity represents forgone federal tax revenue. However, the equity
gained through the sale or offering of tax credits likely would not be
contributed to projects funded under these programs in the absence of
the credit, and therefore may be considered leveraged funds.
[46] The Internal Revenue Code provides broad guidance to states for
allocating tax credit awards, requiring them to consider, among other
things, the extent of a project's financing gap, or the difference
between the cost of a project and the amount of nontax credit financing
that a project can raise to cover those development costs (that is,
leveraged funds). See 26 U.S.C. 42(m)(2).
[47] See 71 Federal Register, 70835, 70841 (Dec. 6, 2006).
[48] HUD distinguishes between matched funds and leveraged funds in the
HOME and HOPE VI programs. Thus, for purposes of our discussion on
performance measurement, we consider match and leverage to be distinct
activities in these programs. As discussed later in this report, HUD
has not identified leveraging as a performance measure for CDBG or HOME
programs.
[49] Extremely-low-income households earn 30 percent or less of the
area median income; very-low-income households earn 50 percent or less;
and low-income households earn 80 percent or less. Although income
limits vary by location, all residents of public housing must be at
least low income.
[50] See GAO-07-768R.
[51] Senior debt must be repaid before subordinated debt receives any
payment in the event of default.
[52] PART asks a series of questions about a program's performance and
management; OMB assigns programs an overall rating--effective,
moderately effective, adequate, ineffective, or results not
demonstrated--based on the results of its PART reviews. OMB conducted
PART reviews of the three HUD programs in 2003.
[53] In its 2003 PART review of the CDBG program, OMB rated the program
"ineffective."
[54] HUD started collecting leveraging data for the program and plans
to publish measures on its Web site by the end of calendar year 2007.
HUD does not plan to cite leverage goals or measures in any of its
performance and budget documents.
[55] Specifically, we found that program design has implications for
the availability of performance information. Among the programs
reviewed, relatively few collected uniform data on outcomes of state or
local activities. Collecting such data requires conditions--uniformity
of activities, objectives, and measures--that do not always exist under
many flexible program designs. See GAO/HEHS/GGD-98-137 and GAO, Grant
Programs: Design Features Shape Flexibility, Accountability, and
Performance Information, GAO/GGD-98-137 (Washington, D.C.: June 22,
1998).
[56] OMB conducted PART reviews of the CDFI program and the Internal
Revenue Service's administration of the New Markets Tax Credit program
in 2004. To date, OMB has not conducted a PART review of the Low-Income
Housing Tax Credit program.
[57] See GAO, Performance Budgeting: Observations on the Use of OMB's
Program Assessment Rating Tool for the Fiscal Year 2004 Budget, GAO-04-
174 (Washington, D.C.: Jan. 30, 2004).
[58] HUD does not publish a leverage measure for the CDBG program.
[59] Treasury does not collect or report leverage data for the Low-
Income Housing Tax Credit program.
[60] See The Community Development Financial Institutions Fund, U.S.
Department of the Treasury, "CDFIs Leverage CDFI Program Awards Nearly
$27 to $1" (Washington, D.C., Feb. 13, 2007) and The Community
Development Financial Institutions Fund, U.S. Department of the
Treasury, "CDFIs Leverage CDFI Program Awards Nearly $20 to $1!"
(Washington, D.C., Mar. 2005).
[61] See Community Development Financial Institutions Fund, "Growth,
Diversity, Impact: A Snapshot of CDFIs in FY 2003" (Washington, D.C.,
June 1, 2007).
[62] As discussed throughout this report, Treasury currently does not
publish a leverage measure for the New Markets Tax Credit program.
[63] HUD maintains the National Low-Income Housing Tax Credit database
that collects information on different funding sources used in tax
credit projects (such as tax-exempt bond financing, Rural Housing
Service Section 515 or Federal Housing Administration loans, and HUD
funds), but not the dollar amounts of these sources. As a result, we
did not use the data in HUD's database to assess the extent of
leveraging in the program.
[64] The CDBG program provides formula-based grants to metropolitan
cities and urban counties, known as entitlement communities, and to
states for distribution to nonentitlement communities, which may carry
out activities directly or award funds to subrecipients. Similarly, the
HOME program provides formula-based grants to states and localities
(certain cities, counties, or consortiums of cities and counties),
which can administer these grants on their own, or with or through
third parties or subgrantees.
[65] HUD uses data from fiscal year 1992 through the most recent month
to calculate the leverage measure it publishes for the HOME program.
However, because HUD officials raised questions about the quality of
the pre-2004 HOME data, we used fiscal year 2006 data in our
calculations of the program's leverage measure.
[66] The HOPE VI program provides grants to public housing agencies to
replace severely distressed public housing units with attractive,
economically viable communities that often combine public housing with
other affordable or market-priced housing units. HUD's HOPE VI data
collection contract expired on Mar. 31, 2006, due to the delayed
approval of the agency's technical assistance plan. Since that date,
HUD has not been able to collect subsequent quarters' data in its
online database.
[67] HUD's reported leverage measure is based on all leveraging that
occurred between Oct. 1, 2005, and May 31, 2006, and does not
necessarily include all phases of all development projects. Because
substantial leveraging can occur in a development from the time HUD
makes a HOPE VI grant to the time all phases of construction are
completed, we determined that data on completed projects would more
accurately reflect the extent of leveraging that occurred in the
program.
[68] Treasury uses CDFIs' audited financial data (including net assets
and liabilities) to calculate the extent of leveraging in the program.
[69] Fifty-nine agencies receive a Low-Income Housing Tax Credit
allocation; however, we excluded 2, suballocating agencies (the
California Housing Finance Agency and the Massachusetts Housing Finance
Agency) because officials of these agencies told us to contact the
primary Low-Income Housing Tax Credit allocating agency in their state.
A suballocating agency receives a portion of its state's Low-Income
Housing Tax Credit allocation from the primary allocating agency in the
state.
[70] We did not review available data to determine their reliability
and calculate an estimated leverage measure for the program.
[71] The background section of this report provides information on the
purpose and scope of these programs.
[72] We use HUD's CDBG and HOME leverage ratios for illustrative
purposes to demonstrate measurement alternatives; however, the values
presented in the related figures should not be construed as
representing actual leveraging. We discuss the limitations of the
leveraging data the agencies and we used to calculate these leverage
measures earlier in this report. App. I contains more detailed
descriptions of our assessments of the reliability of the agencies'
leveraging data.
[73] The program leverage measure HUD publishes on its Web site is
cumulative from fiscal year 1992 through the most recent month; thus,
it may differ slightly from the leverage measure presented in this
report. We used fiscal year 2006 data to calculate this measure
because, according to HUD officials, fiscal year 2003 and earlier data
are not as complete as later data. Further, while HUD's and our program
leverage measures account for all completed HOME activities (rental and
homeownership), the measures for states and localities on HUD's Web
site only account for completed rental activities.
[74] HUD's reported leverage measure (see table 4) is based on all
leveraging that occurred between Oct. 1, 2005, and May 31, 2006, rather
than on data for completed developments or completed phases of
developments. Because substantial leveraging can occur in a development
from the time HUD makes a HOPE VI grant to the time all phases of
construction are completed, we determined that data on completed
projects would more accurately reflect the extent of leveraging that
occurred in the program. Therefore, we used data on completed projects
for purposes of this analysis. HUD's HOPE VI data collection contract
expired on Mar. 31, 2006--as of that date, HUD stopped collecting HOPE
VI data in its online Grant Management Reporting System Prototype
database. Therefore, our calculation includes HOPE VI developments
completed as of Mar. 31, 2006.
[75] At the institutional level, an organization (such as a group of
investors or a community or other development authority) pools funds
from multiple sources, which then are used to finance a portfolio of
projects. At the project level, an organization (such as a state or
local agency) leverages funds as necessary for discrete projects.
[76] The background section and app. II provide more information on how
leveraging occurs in the CDFI program.
[77] Treasury has reported that project leverage in the CDFI program
can vary considerably by loan purpose--for example, in fiscal year 2003
the highest median project leverage ratios were for multifamily housing
rehabilitation (8.53 to 1) and construction (5.35 to 1), while the
lowest were for business loans (between 0.60 to 1 and 0.81 to 1). See
The Community Development Financial Institutions Fund, U.S. Department
of the Treasury, CDFIs Leverage CDFI Program Awards Nearly $20 to $1!
(Washington, D.C., Mar. 2005).
[78] The calculations of institutional and total leverage are based on
the after-tax value of the credit ($0.25). Using the nominal cost of
the credit ($0.39) as the other selected programs generally used,
institutional leverage would be 1.56 to 1 and total leverage would be
3.84 to 1. The background section and app. II provide more information
on how leveraging occurs in the New Markets Tax Credit program.
[79] The background section of this report provides more detailed
information on the selected programs.
[80] While initially the tax credit would appear to represent lost tax
revenue to the Federal government of 39 cents for each dollar invested,
the cost basis of the investment is reduced by the amount of the
credits claimed. Therefore, the investor has to pay taxes (generally at
a corporate tax rate of 35%) on the 39 cents of credits claimed, which
reduces the cost of the credit to the Federal government from 39 cents
to approximately 25 cents [.39 * (1-.35) = .2535].
*** End of document. ***