GAO-23-105952: Community Development Financial Institutions: Analysis of Annual and Emergency Assistance and Treasury Efforts to Address Program Risks
Summary
GAO report GAO-23-105952, "Community Development Financial Institutions: Analysis of Annual and Emergency Assistance and Treasury Efforts to Address Program Risks," a report to the Ranking Member of the Committee on Banking, Housing, and Urban Affairs, U.S. Senate, dated December 2022. It states that the Consolidated Appropriations Act, 2021 appropriated $188 million to Treasury's CDFI Program and supplemental programs and $12 billion in emergency assistance for three new programs. The table shows the Rapid Response Program ($1.25 billion, fully disbursed), the Emergency Capital Investment Program ($9.00 billion, with $8.28 billion disbursed as of Sept. 21, 2022) and the Equitable Recovery Program ($1.75 billion, not yet awarded). GAO found awards from the first two exceeded typical CDFI Program awards (median $400,000 in fiscal year 2021).
Summary drafted by a model from the document's text below and checked by script against that text before publication. It is a navigation aid, not a reading of what the document proves. Where AI is used
Cite this document
“GAO-23-105952: Community Development Financial Institutions: Analysis of Annual and Emergency Assistance and Treasury Efforts to Address Program Risks.”
Original at www.gao.gov. Archived at Pandemic Darlings, https://pandemicdarlings.com/source-documents/gao-23-105952-community-development-financial-institutions-analysis-of-d4fde17275/
, SHA-256 e4484eafc92ac5cbce7adb60ac4439d2df6bc7447dc52ae4a5be6006754e73ce.Full text
United States Government Accountability Office
Report to the Ranking Member,
Committee on Banking, Housing, and
Urban Affairs, U.S. Senate
COMMUNITY
December 2022
DEVELOPMENT
FINANCIAL
INSTITUTIONS
Analysis of Annual
and Emergency
Assistance and
Treasury Efforts to
Address Program
Risks
GAO-23-105952
December 2022
COMMUNITY DEVELOPMENT FINANCIAL
INSTITUTIONS
Analysis of Annual and Emergency Assistance and
Highlights of GAO-23-105952, a report to the Treasury Efforts to Address Program Risks
Ranking Member, Committee on Banking,
Housing, and Urban Affairs, U.S. Senate
Why GAO Did This Study What GAO Found
Community development financial institutions (CDFI) and minority depository
Access to credit and capital is vital to institutions (MDI) provide financial services to underserved communities. The
economic development but is limited in Department of the Treasury assists eligible institutions through annual CDFI
many low- and moderate-income Program awards. Treasury received additional funds in 2021 for three COVID-19-
communities underserved by traditional related assistance programs—Rapid Response Program, Emergency Capital
financial institutions. The economic
Investment Program, and Equitable Recovery Program (see table). GAO found
effects of the COVID-19 pandemic
have further stressed these that awards for the first two of these programs were larger than typical CDFI
communities, which often rely on Program awards, which had a median of $400,000 in fiscal year 2021.
CDFIs and MDIs—generally smaller,
Status of Awards from Treasury’s Emergency Assistance Programs for Community
community-based organizations—for Development Financial Institutions and Minority Depository Institutions
financial products and services.
Program Amount Status Primary recipients Median
The Consolidated Appropriations Act, appropriated (percentage of all recipients) award
2021, appropriated $188 million to Rapid Response $1.25 billion Fully disbursed Loan funds (54%); credit $1.83
Treasury’s existing CDFI Program and Program unions (28%) million
supplemental programs. It also Emergency Capital $9.00 billion $8.28 billion Banks, savings associations, $29.94
appropriated $12 billion in emergency Investment disbursed as of and holding companies (57%); million
assistance for three new Treasury Program Sept. 21, 2022 credit unions (43%)
programs designed to help CDFIs and Equitable Recovery $1.75 billion Not yet awarded Not applicable Not
Program applicable
MDIs respond to the impacts of
COVID-19, most of which has been Source: GAO and Department of the Treasury. | GAO-23-105952
awarded. Two of the new programs GAO’s analysis of CDFI Program and Rapid Response Program data found that
required Treasury to award funds recipients have used awards mainly to provide financial products and to increase
expeditiously.
reserves. Financial products provided by CDFI Program recipients in fiscal years
GAO was asked to assess Treasury’s 2016–2020 mainly funded home purchases, businesses, and microenterprises.
distribution of funds, efforts to ensure
program integrity, and the use of Treasury took steps to address potential risks to program integrity created by
Treasury’s funding. This report
examines (1) the types of institutions aspects of the emergency programs. For example:
that received funds from the CDFI
• Risk-based application reviews. Treasury allowed higher-risk institutions to
Program or the 2021 emergency
programs and how much they be eligible for the Rapid Response Program and the Equitable Recovery
received, (2) how recipients have used Program to be inclusive of institutions from more communities (e.g., those
CDFI Program and Rapid Response with low levels of bank deposits). To address potential risks, Treasury’s
Program funds, and (3) steps Treasury process called for additional reviews of such applicants.
took to address risks posed by the
emergency programs. • Ability-to-pay requirement. The Emergency Capital Investment Program
To conduct this work, GAO analyzed requires recipients to pay Treasury dividends or interest in exchange for a
Treasury data, reviewed Treasury capital investment. To help ensure recipients can pay the amounts due,
policies and procedures, and Treasury required institutions to have at least a 1-to-1 ratio of net income to
interviewed Treasury officials. the dividend or interest amount to be eligible for an investment.
Treasury is also developing a risk-based process for reviewing Rapid Response
Program and Equitable Recovery Program recipients and is finalizing policies for
View GAO-23-105952. For more information,
contact William B. Shear at (202) 512-8678 or post-award reporting and award monitoring for the Emergency Capital
ShearW@gao.gov. Investment Program. These efforts are intended to monitor whether the funds are
being used as intended.
United States Government Accountability Office
Contents
Letter 1
Background 3
Banks and Other Institutions Received Emergency Program
Awards That Were Larger than Annual CDFI Program Awards 8
Recipients Used CDFI Program Funds Largely for Financial
Products and Reserves, and Availability of Borrower Data
Varies 14
Treasury Responded to Potential Risks When Creating
Emergency Programs 30
Agency Comments 41
Appendix I Available Research on Economic Impact of Programs
and Geographic Analysis of CDFI Program Recipients 43
Appendix II Objectives, Scope, and Methodology 49
Appendix III Information on the Lines of Business and
Services Provided by CDFI Program Recipients 55
Appendix IV GAO Contact and Staff Acknowledgments 57
Tables
Table 1. Information on Selected Treasury Programs for
Community Development Financial Institutions (CDFI) and
Minority Depository Institutions (MDI) 5
Table 2: Number and Types of Organizations with Available Data
That Received CDFI Program Awards, Fiscal Years
2019–2021 9
Table 3: Assets, Revenue, and Full-Time Equivalent Staff for
Fiscal Year 2019–2021 CDFI Program Recipients with
Available Data 10
Table 4: Number and Types of Organizations That Received
Rapid Response Program Awards 11
Table 5: Number and Types of Organizations That Received
Emergency Capital Investment Program Awards 11
Page i GAO-23-105952 Community Development Financial Institutions
Table 6: Emergency Capital Investment Program Recipient Assets 12
Table 7: Award Amounts for CDFI Program (Fiscal Years 2019–
2021), Rapid Response Program (RRP), and Emergency
Capital Investment Program (ECIP) 12
Table 8: Number of CDFIs that Received a Combination of Fiscal
Year 2021 CDFI Program, Rapid Response Program
(RRP), or Emergency Capital Investment Program (ECIP)
Awards 13
Table 9: Number of Transactions Made by Institutions Receiving
Financial Assistance from the CDFI Program, Fiscal
Years 2016–2020 20
Table 10: Dollar Amount of Transactions Made by Institutions
Receiving Financial Assistance from the CDFI Program,
by Fiscal Year, 2016–2020 21
Table 11: Types of Borrowers Served by Institutions Receiving
Financial Assistance from the CDFI Program, Fiscal
Years 2016–2020 22
Table 12: Total Dollar Amount of Transactions Made by
Institutions Receiving Financial Assistance from the CDFI
Program, by Transaction Type, Fiscal Years 2016–2020 23
Table 13: Total Dollar Amount of Transactions Made by
Institutions Receiving Financial Assistance from the CDFI
Program, by Purpose, Fiscal Years 2016–2020 27
Table 14: Income Status of Individual Borrowers Served by
Institutions Receiving Financial Assistance from the CDFI
Program, Fiscal Years 2018–2020 28
Table 15: Credit Scores of Individual Borrowers Served by
Institutions Receiving Financial Assistance from the CDFI
Program, Fiscal Years 2018–2020 29
Table 16: Location Type of Community Development Financial
Institutions (CDFI) Program Recipients, Fiscal Years
2019–2021 45
Table 17: Average Socioeconomic and Demographic
Characteristics of Counties in Which Community
Development Financial Institutions (CDFI) Program
Recipients Were and Were Not Located, Fiscal Years
2019–2021 46
Table 18: Average Socioeconomic and Demographic
Characteristics of Census Tracts Where Community
Development Financial Institutions (CDFI) Did or Did Not
Conduct Projects, Fiscal Years 2016–2020 48
Page ii GAO-23-105952 Community Development Financial Institutions
Table 19: Primary Business Lines of CDFI Program Recipients
with Available Data, Fiscal Years 2019–2021 55
Table 20: Types of Development Services CDFI Program
Recipients Offered with Available Data, Fiscal Years
2019–2021 56
Table 21: Types of Financial Services CDFI Program Recipients
Provided with Available Data, Fiscal Years 2019–2021 56
Figures
Figure 1: Uses of Financial Assistance from the CDFI Program
(Fiscal Year 2016 to Reports Received by June 30, 2022)
and Rapid Response Program (RRP) (Reports Received
by June 30, 2022) 16
Figure 2: Uses of Technical Assistance from the CDFI Program
(Fiscal Year 2016 to Reports Received by June 30, 2022)
and Rapid Response Program (RRP) (Reports Received
by June 30, 2022) 18
Figure 3: Types of Transactions Made by Institutions Receiving
Financial Assistance from the CDFI Program, Fiscal
Years 2016–2020 23
Figure 4: Purpose of Transactions Made by Institutions Receiving
Financial Assistance from the CDFI Program, Fiscal
Years 2016–2020 25
Figure 5: Distribution of Community Development Financial
Institution Program Awards by State, Fiscal Years 2016–
2021 47
Page iii GAO-23-105952 Community Development Financial Institutions
Abbreviations
ACR Annual Certification and Data Collection Report
AMFI area median family income
AMIS Awards Management Information System
CAMEL capital adequacy, asset quality, management, earnings,
liquidity/asset-liability management
CAMELS capital adequacy, asset quality, management, earnings,
liquidity, sensitivity to market risk
CDFI community development financial institution
Census ACS Bureau of the Census American Community Survey
ECIP Emergency Capital Investment Program
ERP Equitable Recovery Program
MDI minority depository institution
OIG Office of Inspector General
RRP Rapid Response Program
TLR Transaction Level Report
Treasury Department of the Treasury
USDA U.S. Department of Agriculture
This is a work of the U.S. government and is not subject to copyright protection in the
United States. The published product may be reproduced and distributed in its entirety
without further permission from GAO. However, because this work may contain
copyrighted images or other material, permission from the copyright holder may be
necessary if you wish to reproduce this material separately.
Page iv GAO-23-105952 Community Development Financial Institutions
Letter
441 G St. N.W.
Washington, DC 20548
December 21, 2022
The Honorable Pat Toomey
Ranking Member
Committee on Banking, Housing, and Urban Affairs
United States Senate
Dear Mr. Toomey:
Access to credit and capital is vital to economic development but is
limited in many low- and moderate-income communities underserved by
traditional financial institutions. These communities often rely on
community development financial institutions (CDFI) and minority
depository institutions (MDI) for financial products and services. CDFIs
include community development banks, credit unions, and nonregulated
institutions such as loan and venture capital funds. They are certified by
the Department of the Treasury and have a primary mission of promoting
community development. MDIs are generally regulated banks or credit
unions primarily owned by one or more minority individuals, or institutions
that serve a predominantly minority community and whose board of
directors and account holders are comprised primarily of minorities. 1
The communities CDFIs and MDIs serve were further stressed by the
economic effects of the COVID-19 pandemic. In response, Congress
enacted provisions in the Consolidated Appropriations Act, 2021, that
provide assistance to CDFIs and MDIs. The act appropriated $188 million
to Treasury’s existing CDFI Program (and supplemental programs) and
$12 billion in emergency assistance for three new Treasury programs
designed to help CDFIs and MDIs respond to the impacts of COVID-19. 2
The CDFI Program, which receives annual appropriations, was created in
1994 and offers competitive financial and technical assistance awards to
qualified CDFIs to support and enhance their ability to meet community
needs. 3 The three emergency programs are the Rapid Response
1Throughout this report, we use the term minority for consistency with language used by
agencies, programs, and statutes within the scope of this report.
2Pub. L. No. 116-260, div. E, 134 Stat. 1182, 1383 (2020); Pub. L. 116-260, div. N, §§
522, 523, 134 Stat. 1182, 2079 (2020).
3Riegle Community Development and Regulatory Improvement Act of 1994, Pub. L. No.
103-325, §§ 101-121, 108 Stat. 2160, 2163.
Page 1 GAO-23-105952 Community Development Financial Institutions
Program (RRP), Equitable Recovery Program (ERP), and Emergency
Capital Investment Program (ECIP). 4 RRP and ECIP required Treasury to
make grants to or capital investments in CDFIs and MDIs on an expedited
basis.
You asked us to examine Treasury’s distribution of funds to CDFIs and
MDIs—including steps to ensure emergency program integrity—and to
evaluate the use and impact of program awards. This report examines (1)
the types of institutions that received funds from the CDFI Program or the
2021 emergency programs and how much they received, (2) how
recipients have used CDFI Program and RRP funds, and (3) steps
Treasury took to address risks posed by the emergency programs.
Additionally, appendix I discusses the extent of and findings from
research on the economic impact of the CDFI Program and emergency
programs, and it presents demographic and socioeconomic analysis of
CDFI Program recipients and project locations.
For objective 1, we analyzed Treasury data on awards and award
recipients under the CDFI Program (for fiscal years 2019–2021), RRP,
and ECIP. 5 For the CDFI Program, we began with fiscal year 2019 data
because certain earlier data were unreliable for the analysis we
performed. For objective 2, we analyzed data that CDFI Program
recipients reported to Treasury from fiscal year 2016 through June 30,
2022, and data that RRP recipients reported from the program’s inception
through June 30, 2022, on the general categories of activities they funded
with their awards. Additionally, for the CDFI Program, we analyzed
Treasury data on all of the loans and investments recipients made in
fiscal years 2016–2020, regardless of funding source. To assess the
reliability of Treasury’s data, we reviewed Treasury documentation and
conducted electronic testing, including checks for outliers, missing data
fields, and erroneous values. We determined these data were sufficiently
reliable for purposes of characterizing program recipients, award
amounts, and uses of funds.
For objective 3, we reviewed Treasury’s policies and procedures for the
emergency programs. We identified potential program risks by examining
4Treasury refers to RRP as CDFI RRP and to ERP as CDFI ERP. Treasury also refers to
ECIP award recipients as ECIP participants. For ease of presentation, we shorten the
program abbreviations and use the term recipient for all of the programs.
5Fiscal year 2021 was the most recent year for which CDFI Program data were available
when we conducted our analysis. The data for RRP and ECIP cover awards from funds
appropriated in fiscal year 2021, the only year in which those programs received funds.
Page 2 GAO-23-105952 Community Development Financial Institutions
program eligibility and targeting requirements, recipient selection
processes, and post-award financial terms and reporting requirements.
We also compared the policies and procedures for the emergency
programs to those for the CDFI Program to identify streamlined
processes that could introduce risks to program integrity. We reviewed
agency documentation and interviewed Treasury officials to determine
whether and how Treasury addressed the risks we identified. Additionally,
we reviewed documentation describing Treasury mechanisms to help
prevent fraud, waste, and abuse in the emergency programs. Because
the emergency programs were in the early stages of implementation, we
did not evaluate the extent to which Treasury’s approach aligned with
leading practices for fraud risk management. We also reviewed
information on Treasury’s assessment of improper payment risks. Finally,
we reviewed Treasury Office of Inspector General (OIG) reports on the
emergency programs and Treasury documentation of steps taken to
address OIG recommendations. Appendix II describes our scope and
methodology in greater detail.
We conducted this performance audit from March 2022 to December
2022 in accordance with generally accepted government auditing
standards. Those standards require that we plan and perform the audit to
obtain sufficient, appropriate evidence to provide a reasonable basis for
our findings and conclusions based on our audit objectives. We believe
that the evidence obtained provides a reasonable basis for our findings
and conclusions based on our audit objectives.
Background
CDFIs and MDIs CDFIs are financial institutions that seek to expand economic opportunity
in low-income communities by providing access to financial products and
services for local residents and businesses that otherwise may not be
accessible. For example, they may provide mortgage financing for low-
income and first-time homebuyers or financing for not-for-profit affordable
housing developers. Treasury’s CDFI Fund certifies CDFIs that meet
Page 3 GAO-23-105952 Community Development Financial Institutions
seven statutory and regulatory criteria. 6 CDFIs can be regulated
institutions (such as banks) or nonregulated institutions (such as venture
capital funds). They can also be for-profit or nonprofit institutions and can
be funded by private and public sources. Institutions generally must be
certified CDFIs to be considered for CDFI Fund program awards. 7
MDIs generally are regulated banks, savings associations, or credit
unions primarily owned by one or more minority individuals, or that serve
a predominantly minority community and whose board of directors and
account holders are comprised primarily of minorities. Like CDFIs, MDIs
typically serve economically challenged communities traditionally
underserved by the banking industry and other businesses.
CDFI Program and The negative effects of the COVID-19 pandemic—including job losses
Emergency Programs and business closures—were felt especially hard in lower-income
communities in which CDFIs and MDIs often operate. In the Consolidated
Appropriations Act, 2021, Congress provided annual appropriations for
the existing CDFI Program and emergency funding for new programs
aimed at helping CDFIs and MDIs respond to the impact of COVID-19. 8
As shown in table 1, Congress appropriated $188 million for the CDFI
Program and supplemental programs, and divided $12 billion in
emergency assistance among three new programs administered by
Treasury. Specifically, Congress appropriated $1.25 billion for RRP,
$1.75 billion for ERP, and $9 billion for ECIP. The CDFI Program, RRP,
and ERP are administered by Treasury’s CDFI Fund, while ECIP is
6The CDFI Fund was established by the Riegle Community Development and Regulatory
Improvement Act of 1994, Pub. L. No. 103-325, § 104, 108 Stat. 2160, 2166. To be
certified by the CDFI Fund, an institution must (1) be a legal entity; (2) have a primary
mission of promoting community development; (3) serve principally an investment area or
targeted population; (4) be an insured depository institution, or make loans or
development investments as its predominant business activity; (5) provide development
services (such as technical assistance or counseling) in conjunction with its financing
activity; (6) maintain accountability to its target market; and (7) be a nongovernmental
entity.
7Two types of entities that are not yet certified CDFIs are eligible to receive technical
assistance awards through the CDFI Program or Native American CDFI Assistance
Program: (1) Emerging CDFIs are noncertified entities that demonstrate in their award
applications that they have an acceptable plan to meet CDFI certification requirements by
the end of the award period of performance or another date that the CDFI Fund selects.
(2) Sponsoring entities are legal organizations that primarily serve Native communities
and plan to create a separate legal entity that will become a certified CDFI primarily
serving Native communities by a date specified in the award assistance agreement.
8Pub. L. No. 116-260, 134 Stat. 1182, 1383 (2020); Pub. L. 116-260, div. N, §§ 522, 523,
134 Stat. 1182, 2079 (2020).
Page 4 GAO-23-105952 Community Development Financial Institutions
administered by the Office of Community and Economic Development
within Treasury’s Office of Financial Institutions.
Table 1. Information on Selected Treasury Programs for Community Development Financial Institutions (CDFI) and Minority
Depository Institutions (MDI)
Fiscal year
2021 Award Eligible
Name appropriation type entities Purpose
CDFI Program $188 milliona Loans, grants, Certified CDFIs and Provides financial and technical assistance to CDFIs to
and equity certain emerging invest in and build their capacity and to drive community
supplemental investments, CDFIs with a plan to revitalization.
programs credit union become certified
shares
Rapid $1.25 billion Grants Certified CDFIs Provide grants to support CDFIs in their response to the
Response economic impact of the COVID-19 pandemic.
Program
Equitable $1.75 billion Grants Certified CDFIs Provides grants to expand lending, grant making, and
Recovery doing business in investment activity in low- or moderate-income
Program certain geographic communities and to borrowers, including minorities, that
areas have significant unmet capital or financial services
needs and who were disproportionately impacted by the
COVID-19 pandemic. Restricted to CDFIs with a track
record of working in areas disproportionately affected by
COVID-19.b
Emergency $9 billion Capital Certified CDFIs or Provides direct and indirect capital investments to
Capital investments MDIs support efforts to provide loans, grants, and forbearance
Investment for small and minority-owned businesses and
Program consumers, especially in underserved communities that
may be disproportionately impacted by the COVID-19
pandemic.
Source: GAO analysis. | GAO-23-105952
a
The $188 million includes $165 million for the CDFI Program and two of its supplemental programs,
Persistent Poverty Counties and Disability Funds, and $23 million for another supplemental program,
the Healthy Food Financing Initiative. This amount does not include funding provided for the Native
American CDFI Assistance Program, which is similar to the CDFI Program and received $16.5
million.
We use the term minority for consistency with agency, program, and statute language.
b
Community Development Financial Institutions Program. The CDFI
program provides financial assistance and technical assistance awards
that CDFIs may use for a number of designated activities. The awards
can take several forms, including grants and loans. Financial assistance
awards support CDFI activities, such as offering financial products,
financial services, and development services, or supplementing loan loss
reserves or capital reserves. Most financial assistance awards require
institutions to match the amounts with funds from a source other than the
Page 5 GAO-23-105952 Community Development Financial Institutions
federal government. 9 In addition to base financial assistance, the CDFI
Program has supplemental programs that provide financial assistance for
specific purposes, such as the Healthy Food Financing Initiative. 10
Institutions have 3 full fiscal years in which to use the funds.
Technical assistance awards are grants that support capacity-building
activities of CDFIs, such as compensating staff, purchasing equipment,
and providing training to staff. Emerging CDFIs that are not yet certified
by the CDFI Fund are eligible for certain technical assistance awards.
CDFIs must use the funds within 2 full fiscal years of receipt.
Through the CDFI Fund’s Native American CDFI Assistance Program,
financial assistance and technical assistance awards are available to
CDFIs, emerging CDFIs, and sponsoring entities that seek to increase
access to financial products and services in Native Communities and
create and expand Native CDFIs. 11
Rapid Response Program. RRP provides grants that CDFIs can use for
both financial and capacity-building activities. 12 The Consolidated
Appropriations Act, 2021, required Treasury to make these funds
available within 60 days of enactment. As of June 15, 2021, Treasury had
fully disbursed its appropriation of $1.25 billion. CDFIs must use the funds
within 2 full fiscal years after the award announcement.
Equitable Recovery Program. Like RRP, ERP provides grants that
CDFIs may use for both financial and capacity-building activities, but
9Native CDFIs receiving awards through the CDFI Program and the Native American
CDFI Assistance Program are exempt from the matching funds requirement. The
Consolidated Appropriations Act, 2021, also waived the matching funds requirement for
small and emerging CDFIs and for CDFIs receiving awards through the Healthy Food
Financing Initiative.
10The Healthy Food Financing Initiative provides flexible financial assistance and
specialized training and technical assistance to CDFIs that invest in a range of retail and
nonretail healthy food projects serving low-income communities, including food
production, grocery stores, mobile food retailers, farmers markets, cooperatives, corner
stores, bodegas, and others.
11Native CDFI is the official term Treasury uses to refer to CDFIs that help Native
American communities increase their access to credit, capital, and financial services. To
qualify as a Native CDFI, 50 percent or more of an institution’s activities must be directed
to a Native community.
12Recipients may spend the greater of $200,000 or 15 percent of the grant on capacity-
building activities.
Page 6 GAO-23-105952 Community Development Financial Institutions
eligibility is limited to certified CDFIs that provide services in areas of the
country that Treasury designated as disproportionately affected by
COVID-19. 13 The application period for awards closed on September 22,
2022, and as of October 2022, Treasury was reviewing applications.
CDFIs will have 5 fiscal years in which to spend the funds.
Emergency Capital Investment Program. ECIP provides long–term
capital investments in federally insured depository institutions, bank or
savings and loan holding companies, and federally insured credit unions
in exchange for preferred stock or subordinated debt issued to Treasury.
Recipients pay Treasury dividends or interest for the period of the
investment, beginning after the first 2 years. The maximum dividend or
interest rate is 2 percent of Treasury’s investment per year, but this
amount may be lowered if the participating institution increases its lending
to certain target markets above a baseline. 14 The rate is recalculated
each year until the 10th year after the investment, at which point the rate
is fixed. The Consolidated Appropriations Act, 2021, required Treasury to
begin accepting applications within 30 days of enactment. Treasury
began accepting applications on March 4, 2021, a little more than a
month past the January 26, 2021, deadline. As of September 21, 2022,
13Recipients may spend the greater of $166,667 or 15 percent of the grant, up to
$400,000, on capacity-building activities. Additionally, there are restrictions on the amount
of the award that may be used on certain financial activities. Recipients may also use
funds to provide grants.
14Specifically, the dividend or interest rate will be 1.25 percent of Treasury’s investment
per year if the institution increases its qualified lending by 200–400 percent of the
investment above the baseline, or 0.5 percent of Treasury’s investment per year if the
institution increases its qualified lending by more than 400 percent of the investment
above the baseline. The baseline is the dollar amount of qualified lending the institution
originated during the 12-month period ending September 30, 2020.
Page 7 GAO-23-105952 Community Development Financial Institutions
Treasury provided $8.28 billion in investments. ECIP investments were
available to both CDFIs and MDIs that were not certified as CDFIs. 15
Banks and Other
Institutions Received
Emergency Program
Awards That Were
Larger than Annual
CDFI Program
Awards
Recipient Institutions Were For the period we examined, CDFI Program, RRP, and ECIP recipients
Mostly Loan Funds, were mostly loan funds, credit unions, and banks, although the proportion
of organization types varied by program. 16 Variation among the programs
Banks, and Credit Unions
is partially attributable to the different program eligibility criteria and
of Varying Sizes purposes. Our analysis covered institutions that received fiscal year
2019–2021 awards from the CDFI Program and institutions that received
RRP and ECIP awards.
CDFI Program recipients. According to available Treasury certification
information, CDFI Program recipients from fiscal years 2019 through
15For ECIP, an MDI is an institution that is: (1) A “minority depository institution,” as
defined in section 308 of the Financial Institutions Reform, Recovery, and Enforcement
Act of 1989 (12 U.S.C. 1463 note). A depository institution satisfies this definition if it is (a)
a privately owned institution, 51 percent of which is owned by one or more socially and
economically disadvantaged individuals; (b) a publicly owned institution, 51 percent of the
stock of which is owned by one or more socially and economically disadvantaged
individuals; or (c) a mutual institution of which the majority of the board of directors,
account holders, and the community which it services is predominantly minority (defined
by statute as Black American, Native American, Hispanic American, or Asian American).
(2) An institution considered to be an MDI by the Office of the Comptroller of the Currency,
the Federal Deposit Insurance Corporation, the Board of Governors of the Federal
Reserve System, or the National Credit Union Administration, as applicable.
16For data analysis purposes, we use the term CDFI Program to include institutions that
received financial or technical assistance awards from the CDFI Program or the Native
American CDFI Assistance Program, as well as those that received supplemental
assistance awards for the Healthy Food Financing Initiative or Persistent Poverty
Counties.
Page 8 GAO-23-105952 Community Development Financial Institutions
2021 were mostly loan funds and credit unions (see table 2). 17 Given data
availability limitations, our analysis included 503 recipients, or about 68
percent. For all years combined, about 62 percent of these 503 recipients
were loan funds, and about 24 percent were credit unions. Banks, thrifts,
and bank holding companies accounted for about 13 percent of
recipients, while venture capital funds represented about 1 percent.
Table 2: Number and Types of Organizations with Available Data That Received CDFI Program Awards, Fiscal Years 2019–
2021
Bank holding Bank or Credit Loan Venture
company thrift union fund capital fund Total
Number 59 4 122 312 6 503
Percentage 11.7% 0.8% 24.3% 62.0% 1.2% 100%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: For our analysis, we matched Treasury’s fiscal year 2019–2020 certification data to
Community Development Financial Institutions (CDFI) Program awards data for fiscal years 2019–
2021. Because of the unavailability of fiscal year 2021 certification data and because not all CDFIs
who received awards were required to submit these data, this analysis includes the approximately 68
percent of recipients for which data were available.
The data in the table include institutions that received financial or
technical assistance awards from the CDFI Program or the Native
American CDFI Assistance Program, as well as those that received
supplemental assistance awards for the Healthy Food Financing Initiative
or Persistent Poverty Counties.
About 10 percent (48) of the CDFI Program recipients from fiscal years
2019 through 2021 with available data were Native CDFIs. Like the wider
population of award recipients, Native CDFI recipients were largely loan
funds (about 85 percent) or credit unions (about 8 percent).
As shown in table 3, the assets, revenue, and number of full-time
equivalent staff of CDFI Program recipients during fiscal years 2019
through 2021 with available data varied widely. Assets ranged from
$19,670 to about $11 billion. Revenue ranged from $366 to about $527
17Fiscal year 2021 CDFI certification data were not yet available when we conducted our
analysis. For our analysis, we matched Treasury’s fiscal year 2019–2020 certification data
to CDFI Program awards data for fiscal years 2019–2021. Our analysis does not capture
all recipients because of the unavailability of fiscal year 2021 certification data and
because not all CDFIs that received awards in fiscal years 2019–2021 were required to
submit annual certification data (e.g., emerging CDFIs). Our analysis includes the 68
percent of recipients for which data were available.
Page 9 GAO-23-105952 Community Development Financial Institutions
million. Some CDFI Program awards, such as technical assistance
awards, required that institutions have no more than $250 million in total
assets. 18 The number of full-time equivalent staff ranged from 0 to
1,831. 19
Table 3: Assets, Revenue, and Full-Time Equivalent Staff for Fiscal Year 2019–2021 CDFI Program Recipients with Available
Data
Minimum Maximum
Assets $19,670 $10,534,613,558
Revenue $366 $527,105,640
Full-time equivalent staffa 0 1,831
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: For our analysis, we matched Treasury’s fiscal year 2019–2020 certification data to
Community Development Financial Institutions (CDFI) Program awards data for fiscal years 2019–
2021. Because of the unavailability of fiscal year 2021 certification data and because not all CDFIs
who received awards were required to submit these data, this analysis includes the approximately 68
percent of recipients for which data were available.
The data in the table include institutions that received financial or technical assistance awards from
the CDFI Program or the Native American CDFI Assistance Program, as well as those that received
supplemental assistance awards for the Healthy Food Financing Initiative or Persistent Poverty
Counties.
a
It is possible that recipients that reported zero full-time equivalent staff were operated by part-time
employees or volunteers whose total work hours were less than one full-time equivalent, which
Treasury defines as at least 35 hours per week.
For additional information about fiscal year 2019–2021 CDFI Program
recipients, including information on their primary lines of business and
services provided, see appendix III.
Rapid Response Program recipients. The 863 RRP recipients were
primarily loan funds (about 54 percent) or credit unions (about 28
percent), as shown in table 4. Banks and thrifts accounted for 16 percent
of recipients, while other types of organizations each represented 1
percent of recipients or less. RRP recipients were required to be certified
CDFIs. About 6 percent of RRP recipients (56) were Native CDFIs.
18Certain technical assistance applicants must have lower asset amounts. For example,
venture capital funds seeking technical assistance awards must have no more than $5
million in total assets.
19Recipients that reported zero full-time equivalent staff could have been operated by part-
time employees or volunteers whose total work hours were less than one full-time
equivalent, which Treasury defines as at least 35 hours per week.
Page 10 GAO-23-105952 Community Development Financial Institutions
Table 4: Number and Types of Organizations That Received Rapid Response Program Awards
Bank holding Bank Credit Loan Venture
company or thrift union fund capital fund Total
Number 12 137 243 463 8 863
Percentage 1.4% 15.9% 28.2% 53.7% 0.9% 100%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Note: According to Treasury officials, the number of organizations in certain categories differ slightly
from the CDFI Fund’s June 15, 2021 award announcement because of subsequent changes in
recipient institution type or certification status. Percentages do not add to 100 because of rounding.
Emergency Capital Investment Program recipients. All 162 ECIP
recipients that have received investments as of September 21, 2022,
were either banks, savings associations, holding companies, or credit
unions (see table 5). The higher concentration of banks, holding
companies, and credit unions among ECIP recipients compared to RRP
or the CDFI Program reflects ECIP’s narrower eligibility criteria. ECIP
applicants were required to be low- and moderate-income community
financial institutions in good standing with their respective federal
regulators, whereas CDFI Program and RRP applicants were not required
to be federally regulated financial institutions. 20 Most ECIP recipients
were CDFIs (about 68 percent). MDIs comprised about 15 percent of
recipients, and the remainder (about 17 percent) were both CDFIs and
MDIs.
Table 5: Number and Types of Organizations That Received Emergency Capital Investment Program Awards
Bank, savings association, Credit
or holding company union Total
Number 92 70 162
Percentage 56.8% 43.2% 100%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
As shown in table 6, the assets of ECIP recipients covered a broad
range—from about $1 million to about $14 billion—and were about $335
million at the median. Under the Consolidated Appropriations Act, 2021,
the maximum investment each institution could receive was a percentage
of its asset size, with a maximum of $250 million per institution. The dollar
20Eligibility for the CDFI Program and RRP was open to certified CDFIs that were bank
holding companies, banks or thrifts, credit unions, depository institution holding
companies, loan funds, or venture capital funds. Eligibility for ECIP was open to federally
insured depository institutions, bank or savings and loan holding companies, and federally
insured credit unions.
Page 11 GAO-23-105952 Community Development Financial Institutions
value of the instruments recipients proposed to provide to Treasury
usually matched or were greater than the maximum ECIP award for an
institution.
Table 6: Emergency Capital Investment Program Recipient Assets
Minimum 25th percentile Median 75th percentile Maximum
$1,402,073 $150,038,000 $334,518,500 $666,398,000 $13,815,569,290
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Institutions Received Treasury received larger appropriations for the emergency programs,
Large Emergency Awards which were intended to address the effects of COVID-19, compared to
the CDFI Program’s annual appropriations. Partly for this reason, RRP
Compared with Annual
grant awards and ECIP capital investments were higher than historical
CDFI Program Awards CDFI Program awards. Additionally, Treasury officials cited the nature of
ECIP awards—capital investments that may support long-term growth—
as an important distinction in comparing award sizes across programs.
As shown in table 7, the median awards for both RRP and ECIP—about
$1.8 million and $30 million, respectively—were substantially larger than
the median awards historically provided under the CDFI Program. From
fiscal years 2019 through 2021, the median CDFI Program award ranged
from a low of $400,000 (fiscal year 2021) to a high of $565,000 (fiscal
year 2019).
Table 7: Award Amounts for CDFI Program (Fiscal Years 2019–2021), Rapid Response Program (RRP), and Emergency Capital
Investment Program (ECIP)
Fiscal 25th 75th
year Program Minimum percentile Median percentile Maximuma
2019 CDFI Programb $37,500 $217,000 $565,000 $714,000 $2,000,000
2020 CDFI Program $100,000 $125,000 $520,000 $657,000 $2,750,000
2021 CDFI Program $40,000 $125,000 $400,000 $625,000 $4,250,000
RRP $200,000 $1,105,000 $1,826,265 $1,826,265 $1,826,265
ECIP $38,000 $7,375,000 $29,936,500 $70,923,000 $250,000,000
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
a
The CDFI Program data provide information on individual financial assistance, technical assistance,
or Healthy Food Financing Initiative awards from the CDFI Program or the Native American CDFI
Assistance Program. The financial assistance award amounts include any Persistent Poverty
Counties funds received by the institution.
b
The maximums in this column for the CDFI Program reflect those for the CDFI Program’s Healthy
Food Financing Initiative, which provides awards to CDFIs that invest in a range of retail and nonretail
healthy food projects serving low-income communities. The maximums for the Healthy Food
Financing Initiative are generally higher than the CDFI Program’s and Native American CDFI
Page 12 GAO-23-105952 Community Development Financial Institutions
Assistance Program’s maximum award amounts for financial or technical assistance. For example,
the Healthy Food Financing Initiative’s maximum award amount in 2021 was $5 million, whereas the
CDFI Program’s maximum base award for financial assistance was $1 million.
Among the three programs, ECIP had the largest maximum award ($250
million), followed by the CDFI Program ($4.25 million for fiscal year 2021),
and RRP (about $1.8 million). 21 In RRP, the median and maximum
awards were the same, meaning that many applicants received the
maximum award.
ECIP awards varied most widely, reflecting diversity in the size of
institutions that applied for capital investment. As stated previously, a
recipient’s asset size factored into the maximum investment amount
provided to recipients.
Some Institutions In some cases, CDFIs received awards from both the CDFI Program and
Received Both Emergency one or both of the emergency programs (see table 8). Over 60 percent of
the 453 institutions that received fiscal year 2021 CDFI Program awards
and Regular Program
also received an RRP award, which is to be expected given that the
Awards universe of eligible institutions is similar between programs (i.e., certified
CDFIs). About 11 percent of the CDFI Program recipients also received
an ECIP award, and about 6 percent received both RRP and ECIP
awards.
Table 8: Number of CDFIs that Received a Combination of Fiscal Year 2021 CDFI Program, Rapid Response Program (RRP), or
Emergency Capital Investment Program (ECIP) Awards
Programs Number of institutions that received awards
CDFI Program (total) 453
CDFI Program and RRP 292
CDFI Program and ECIP 48
CDFI Program, RRP, and ECIP 27
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Note: The data in the table include institutions that received financial or technical assistance awards
from the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program, as well as those that received supplemental assistance awards for the Healthy
Food Financing Initiative or Persistent Poverty Counties.
21The maximums for the CDFI Program reflect those for the Healthy Food Financing
Initiative—a component of the CDFI Program—which provides awards to CDFIs that
invest in a range of retail and non-retail healthy food projects serving low-income
communities. The maximum award amounts for this component are generally higher than
the CDFI Program’s maximum award amounts for both base financial assistance awards
and technical assistance awards.
Page 13 GAO-23-105952 Community Development Financial Institutions
Among the institutions that received a fiscal year 2021 CDFI Program
award and one or more emergency program awards, the median
emergency program award was larger than the median CDFI Program
award, as follows:
• For institutions that received both CDFI Program and RRP awards,
the median award was $525,000 for the CDFI Program and about
$1.8 million for RRP.
• For institutions that received both CDFI Program and ECIP awards,
the median award was $525,000 for the CDFI Program and about $32
million for ECIP.
• For institutions that received awards from all three programs, the
median award was $590,000 for the CDFI Program, about $1.8 million
for RRP, and about $16 million for ECIP.
Institutions that received funding from the CDFI Program or RRP have
Recipients Used mainly used their awards to provide financial products, like loans and
CDFI Program Funds investments, and to increase their capital reserves and loan loss
reserves. For the CDFI Program specifically, data on all of the financial
Largely for Financial products recipients provided—using program funding or not—show that
Products and these products primarily funded microenterprise, home purchases, and
business purposes. However, some data on the characteristics of the
Reserves, and borrowers of these loans were unavailable.
Availability of
Borrower Data Varies
CDFI Program and Rapid The CDFI Fund collects data from institutions on the general types of
Response Program activities its CDFI Program and RRP awards funded. Institutions report
the dollar amount of the award that they spent on activities in each of a
Awards Have Largely
predefined set of categories. This set of categories differs for financial
Funded Financial Products
and Reserves
Page 14 GAO-23-105952 Community Development Financial Institutions
assistance awards and technical assistance awards. 22 For the CDFI
Program, the analyses reflect spending recipients reported to Treasury
from fiscal year 2016 to June 30, 2022. For RRP, which began in June
2021, the data reflect spending recipients reported from the program’s
inception through June 30, 2022.
Financial assistance. For both the CDFI Program and RRP, institutions
reported spending the majority of their financial assistance awards on
financial products, such as loans and investments (see fig. 1). However,
the share of awards that went toward financial products was larger for the
CDFI Program (about $617 million, or 70 percent) than for RRP (about
$333 million, or 51 percent). Meanwhile, RRP recipients spent a larger
share of their awards on capital reserves (about $242 million, or 37
percent) than CDFI Program recipients did (about $74 million, or 8
percent). CDFI Program recipients spent slightly more of their awards on
loan loss reserves (15 percent, compared to 10 percent for RRP
recipients). 23
22Institutions can receive an award from the CDFI Program for either financial assistance
or technical assistance, but not both, in the same funding round. Financial assistance
awards must be used for activities in six specific categories (such as providing financial
products or services), while technical assistance awards must be used for activities in a
different set of seven categories (such as staff compensation or travel costs). In contrast,
RRP awards can be used for activities in any of these financial or technical assistance
categories. However, there is a limitation on how much of the RRP award can be spent on
the seven categories associated with CDFI Program technical assistance awards
(recipients may spend up to $200,000 or 15 percent of their grant, whichever is greater, on
activities in the seven categories). In this section, we use the term “financial assistance
awards” to refer to both financial assistance awards from the CDFI Program and reported
spending by RRP recipients in the financial assistance categories. Likewise, “technical
assistance awards” refers to both technical assistance awards from the CDFI Program
and reported spending by RRP recipients in the technical assistance categories.
23The capital reserves category includes funding set aside to support recipients’ ability to
leverage other capital, for purposes such as increasing net assets or providing financing.
The loan loss reserves category includes funding set aside to cover losses on loans,
accounts, and notes receivable or for related purposes. For both programs, only regulated
institutions are eligible to spend program funding in the capital reserves category, as well
as the financial services category. For RRP, recipients of awards greater than $200,000
are required to meet certain benchmarks for the amount of money they provide in financial
products during the program’s period of performance. These benchmarks are tied to the
size of the award they receive.
Page 15 GAO-23-105952 Community Development Financial Institutions
Figure 1: Uses of Financial Assistance from the CDFI Program (Fiscal Year 2016 to Reports Received by June 30, 2022) and
Rapid Response Program (RRP) (Reports Received by June 30, 2022)
Note: The graphic on the left includes data from institutions that received financial assistance awards
from the Community Development Financial Institutions (CDFI) Program, but not those that received
awards from the Native American CDFI Assistance Program or the CDFI Program’s supplemental
programs, such as the Healthy Food Financing Initiative.
Additionally, over time, CDFI Program recipients spent a smaller share of
their award dollars on providing financial products (about 77 percent in
fiscal year 2016, compared to 64 percent in fiscal year 2021) and a larger
Page 16 GAO-23-105952 Community Development Financial Institutions
share on bolstering their capital reserves (4 percent in fiscal year 2016
versus 14 percent in fiscal year 2021). 24
For both programs, recipients spent small shares (about 1 to 3 percent) of
their awards on the categories of financial services, such as providing
checking or savings accounts; development services, such as providing
financial or credit counseling, homeownership counseling, or business
planning and management assistance; and direct administrative
expenses.
Technical assistance. As shown in figure 2, both CDFI Program and
RRP recipients reported spending the majority of their technical
assistance awards on staff compensation, although the proportion was
higher for RRP (76 percent, or about $7 million) than for the CDFI
Program (63 percent, or about $22 million). 25 Compared to RRP
recipients, CDFI Program recipients spent somewhat larger shares of
their technical assistance awards on professional services (16 percent
versus 9 percent) and equipment (15 percent versus 11 percent).
Recipients of both programs spent small amounts (about 1 to 2 percent
each) on supplies, training and education, and travel.
24CDFIs report data according to their institution’s own fiscal year, which differs across
institutions. We grouped reports by the calendar year in which they end, regardless of the
exact dates they cover.
25Our analysis combined two categories—salary and personnel services/fringe benefits—
into a single category for staff compensation.
Page 17 GAO-23-105952 Community Development Financial Institutions
Figure 2: Uses of Technical Assistance from the CDFI Program (Fiscal Year 2016 to Reports Received by June 30, 2022) and
Rapid Response Program (RRP) (Reports Received by June 30, 2022)
Note: The graphic on the left includes data from institutions that received technical assistance awards
from the Community Development Financial Institutions (CDFI) Program, but not those that received
awards from the Native American CDFI Assistance Program.
Recipients Made Loans of The CDFI Fund requires institutions receiving financial assistance from
Varying Size and Number, the CDFI Program to submit data annually on all of the transactions (such
and Mostly for
Microenterprise, Home
Purchases, and Business
Purposes
Page 18 GAO-23-105952 Community Development Financial Institutions
as loans and investments) they make. 26 These data do not specify which
transactions resulted directly from CDFI Program awards, but instead
reflect institutions’ lending activity as a whole. However, the data provide
insight into the types of financial products that program dollars helped to
support. 27
As shown in table 9, the number of transactions CDFIs reported making
per year varied widely across institutions, ranging from just one to tens of
thousands. 28 From fiscal years 2016 through 2020, the average (mean)
annual number of transactions per CDFI was 523, while the median
number of transactions was 53. The table shows that the mean and
maximum number of transactions increased sharply between fiscal years
2016 and 2017. This is because of one particular CDFI that submitted
data for fiscal year 2017 and each year thereafter. This CDFI, which was
active in 19 communities as of 2021, made a much greater number of
transactions than any other recipient and is responsible for the maximum
values displayed in the table from fiscal years 2017 through 2020.
26Recipients of RRP awards are also required to submit such data; however, these data
were not yet available when we conducted our analyses. In addition, the data we analyzed
do not include any consumer loans made by banks or credit unions. (Consumer loans are
defined as personal loans made for health, education, emergency, credit repair, debt
consolidation, or other consumer purposes, excluding loans made for home purchase,
home improvement, or business purposes.) This is because banks and credit unions
report such data separately, aggregated at the census tract level, rather than reporting
each individual transaction.
27Treasury officials said they do not ask program recipients to report which transactions
were funded using program dollars. The officials said they do not have an operational
need for this information, given that institutions can achieve the CDFI Program’s
performance goals and measures using any funds on their balance sheets.
28Treasury officials said they follow up with CDFIs that reported fewer than 10
transactions to confirm that the data are complete. They noted that some CDFIs are small
institutions that only make a small number of loans per year.
Page 19 GAO-23-105952 Community Development Financial Institutions
Table 9: Number of Transactions Made by Institutions Receiving Financial Assistance from the CDFI Program, Fiscal Years
2016–2020
Fiscal year Mean Minimum 25th percentile Median 75th percentile Maximum
2016 159 1 15 41 106 4,269
2017a 558 1 15 48 162 82,703
2018 539 1 21 56 229 84,472
2019 559 1 18 51 209 98,453
2020 680 1 19 68 379 81,185
2016–2020 523 1 18 53 200 98,453
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Note: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
a
The reason for the large increase in the number of transactions from 2016 to 2017 is the inclusion of
one particular CDFI that reported a much larger number of transactions than any other institution.
This CDFI submitted data for fiscal years 2017–2020. The number of transactions it reported is
shown in the maximum column for fiscal years 2017–2020. The second-highest number of
transactions reported by any CDFI each year was much smaller—7,730 in fiscal year 2017, 6,723 in
fiscal year 2018, 7,385 in fiscal year 2019, and 10,889 in fiscal year 2020.
The dollar amount of transactions also varied greatly, ranging from $5 to
tens of millions of dollars. 29 However, as table 10 shows, most
transactions were for amounts in the thousands or tens of thousands of
dollars. From fiscal years 2016 through 2020, the average transaction
made by CDFI Program and Native American CDFI Assistance Program
recipients was for about $86,000, while the median was for $7,700. The
table also shows the effect of the CDFI that reported a large number of
transactions beginning in fiscal year 2017. From fiscal years 2017 through
2020, this CDFI’s transactions accounted for between 30 and 60 percent
of the total transactions from all institutions. The transactions reported by
this CDFI were for very small amounts, so its inclusion in the data has the
effect of reducing the mean, 25th percentile, median, and 75th percentile
amounts from fiscal years 2017 through 2020, as compared to fiscal year
2016.
29Treasury officials said they follow up with CDFIs to confirm the accuracy of the data if
the amounts reported are less than $10 or more than $10 million. They noted that many of
the small amounts are correct, since some CDFIs provide loans of small amounts that are
intended to help borrowers build credit. However, in a few cases, CDFIs reported
transactions of $0 or $1. Treasury said these amounts are usually lines of credit for which
the data reporters could not locate the credit limit, so we excluded these 181 transactions
from our analysis.
Page 20 GAO-23-105952 Community Development Financial Institutions
Table 10: Dollar Amount of Transactions Made by Institutions Receiving Financial Assistance from the CDFI Program, by
Fiscal Year, 2016–2020
Fiscal year Mean Minimuma 25th percentile Median 75th percentile Maximum
2016 $104,215 $44 $5,000 $20,906 $80,495 $50.0 million
2017b $63,851 $5 $2,000 $3,500 $21,836 $22.4 million
2018 $80,970 $7 $2,500 $5,500 $58,000 $20.6 million
2019 $83,614 $6 $2,700 $7,000 $70,000 $31.8 million
2020 $98,741 $11 $3,800 $15,970 $96,000 $57.5 million
2016–2020 $85,646 $5 $2,675 $7,700 $66,000 $57.5 million
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Note: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
a
A few CDFIs reported transactions with dollar amounts of $0 or $1. Treasury said these amounts are
usually lines of credit for which the data reporters could not locate the credit limit, so we excluded
these 181 transactions from our analysis. For other transactions with small amounts, Treasury said
that these amounts are correct, noting that some CDFIs provide loans of small amounts that are
intended to help borrowers build credit.
b
The reason for the decrease in the mean, 25th percentile, median, and 75th percentile between
fiscal years 2016 and 2017 is the inclusion of one CDFI that reported a much larger number of
transactions than any other institution. This CDFI submitted data for fiscal years 2017–2020. It
reported many small loans, which had the effect of reducing the dollar amounts shown in the mean,
25th percentile, median, and 75th percentile columns.
For each transaction they made, institutions reported whether the
borrower was an individual, a business, or another certified CDFI. 30 From
fiscal years 2016 through 2020, about two-thirds of transactions went to
businesses, one-third to individuals, and less than 1 percent to CDFIs
(see table 11). The high number of transactions contributed by the CDFI
discussed previously—which only lends to businesses—resulted in a shift
beginning in fiscal year 2017. Businesses made up less than one-half of
all borrowers in fiscal year 2016, compared to about two-thirds or more in
every year thereafter. However, in all 5 years, a majority of transaction
dollars went to businesses, with over $43 billion going to businesses
across the full period, compared to $29 billion to individuals. 31
30Businesses include both for-profit and nonprofit entities, as well as other entities, such
as tribal or village governments.
31The data do not include any consumer loans made by banks or credit unions. Since
consumer loans are loans to individuals, the actual number and dollar amount of
transactions that were made to individuals is greater than these data indicate.
Page 21 GAO-23-105952 Community Development Financial Institutions
Table 11: Types of Borrowers Served by Institutions Receiving Financial Assistance from the CDFI Program, Fiscal Years
2016–2020
Number and percentage of transactions made to each borrower type
Fiscal year Individuals Businesses CDFIs
2016 22,189 (56%) 17,393 (44%) 45 (0.1%)
2017 34,518 (25%) 103,852 (75%) 101 (0.1%)
2018 59,058 (33%) 120,761 (67%) 137 (0.1%)
2019 74,710 (35%) 140,892 (65%) 151 (0.1%)
2020 98,529 (36%) 173,272 (64%) 106 (0.0%)
2016–2020 289,004 (34%) 556,170 (66%) 540 (0.1%)
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
Percentages may not add to 100 because of rounding.
Term loans accounted for 81 percent of all transactions from fiscal years
2016 through 2020, making them the most frequent type of financial
product institutions provided (see fig. 3). Lines of credit comprised
another 8 percent, while products categorized as “other” made up about
11 percent. 32 Equity investments, debt with equity, loan guarantees, loan
loss reserves, and capital reserves comprised less than 1 percent of
transactions collectively. 33
32The CDFI that reported a very large number of transactions contributed many of those
categorized as “other.” When excluding this organization from the data, “other”
transactions accounted for a smaller share (2 percent instead of 11 percent), and lines of
credit made up a larger share (14 percent instead of 8 percent).
33The loan guarantee, loan loss reserve, and capital reserve categories were not available
as answer choices on the data reporting form in fiscal years 2016–2017. The CDFI Fund’s
guidance to institutions defines the transaction types as follows: A term loan is for a
specific amount and has a specified repayment schedule and a fixed or floating interest
rate; an equity investment is money invested in a company by the CDFI in exchange for
the ownership shares of that company; a line of credit is an arrangement that establishes
the maximum loan amount the customer can borrow; debt with equity is a loan with an
agreement stipulating that it may be converted to an equity upon meeting specified
conditions; and a loan guarantee is a loan agreement where the CDFI is a guarantor. The
final two types involve the CDFI’s use of the financial assistance award from Treasury: In
a loan loss reserve, the CDFI uses the award to reserve for expected losses (e.g.,
because of uncollected loan payments), and in a capital reserve, the CDFI uses the award
to leverage the loan.
Page 22 GAO-23-105952 Community Development Financial Institutions
Figure 3: Types of Transactions Made by Institutions Receiving Financial
Assistance from the CDFI Program, Fiscal Years 2016–2020
Notes: The data in the figure include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level. The
following transaction types are not shown in the figure because, together, they accounted for less
than 1 percent of transactions: equity investment, debt with equity, loan guarantee, loan loss reserve,
and capital reserve. The latter three categories were unavailable in fiscal years 2016–2017.
Table 12 provides the total dollar amount of transactions that institutions
made in each category from fiscal years 2016 through 2020, as well as
the percentage of total dollars in each category.
Table 12: Total Dollar Amount of Transactions Made by Institutions Receiving Financial Assistance from the CDFI Program,
by Transaction Type, Fiscal Years 2016–2020
Transaction type Total amount Percentage of total
Term loan $60,548,524,257 83.6%
Line of credit $10,473,446,643 14.5%
Equity investment $132,777,274 0.2%
Debt with equity $8,029,958 0.0%
Loan guarantee $100,211,828 0.1%
Loan loss reserve $23,621,566 0.0%
Capital reserve $91,763,214 0.1%
Other $1,038,386,576 1.4%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Page 23 GAO-23-105952 Community Development Financial Institutions
Notes: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
Percentages do not add to 100 because of rounding.
The most common purpose of the transactions recipients made was
microenterprise, defined as financing to a for-profit or nonprofit enterprise
with five or fewer employees, with an amount no more than $50,000 (and
not for real estate purposes). Microenterprises accounted for just over
one-half (52 percent) of loans overall and nearly 78 percent of loans
made to business borrowers (see fig. 4). This pattern is primarily due to
the high number of transactions contributed by the CDFI discussed
previously. When excluding this organization, microenterprise
transactions accounted for 18 percent of the total.
Page 24 GAO-23-105952 Community Development Financial Institutions
Figure 4: Purpose of Transactions Made by Institutions Receiving Financial Assistance from the CDFI Program, Fiscal Years
2016–2020
Notes: The data in the figure include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
Borrowers may be classified as either individuals, businesses, or other CDFIs. CDFI borrowers are
included in the “All borrowers” chart but are not shown in a separate chart. This is because
transactions to these borrowers accounted for less than 1 percent of all transactions, and 96 percent
of them had a purpose of “other,” consistent with Treasury instructions.
The consumer category does not include transactions made by banks and credit unions, since these
institutions do not report these data together with other transactions (instead, they report them in
aggregate by census tract).
Across all transactions, 18 percent were for home purchases, 10 percent
for business purposes, and 9 percent for home improvement. When
excluding the CDFI that reported a large number of transactions, all of
Page 25 GAO-23-105952 Community Development Financial Institutions
these categories made up a greater share of the total (31 percent, 17
percent, and 15 percent, respectively). Transactions to individuals were
not affected by this CDFI, which only lends to businesses. As shown in
figure 4, most transactions to individuals were for home purchases (54
percent), home improvement (26 percent), or consumer purposes (15
percent). 34
In terms of the dollar volume of transactions, the largest amount, $23.8
billion, was for home purchases (see table 13). The second-largest
amount, $17.1 billion, was for business purposes (excluding real estate
business). Because microenterprise loans tend to be for small amounts,
less than 4 percent of the total dollar amount went toward these
transactions, even though they accounted for over one-half of all
transactions. Conversely, although small percentages of transactions
were for constructing or rehabilitating residential or commercial real
estate, these transactions tended to be for larger amounts and therefore
comprise greater shares of the total dollars.
34As noted above, the data do not include any consumer loans made by banks or credit
unions, so the actual share of transactions that were for consumer purposes is higher than
these data indicate.
Page 26 GAO-23-105952 Community Development Financial Institutions
Table 13: Total Dollar Amount of Transactions Made by Institutions Receiving Financial Assistance from the CDFI Program,
by Purpose, Fiscal Years 2016–2020
Purpose Total amount Percentage of total
Business $17,093,104,526 23.6%
Consumera $113,363,789 0.2%
Home improvement $4,002,763,443 5.5%
Home purchase $23,763,324,295 32.8%
Microenterprise $2,700,012,036 3.7%
Single-family housing construction $2,511,967,373 3.5%
Single-family housing rehabilitation or acquisition $919,786,488 1.3%
Multifamily housing construction $3,537,214,163 4.9%
Multifamily housing rehabilitation or acquisition $3,586,136,105 5.0%
Commercial real estate construction, permanent financing, $9,206,307,346 12.7%
or acquisition without rehabilitation
Commercial real estate rehabilitation $2,376,471,060 3.3%
Other, including transactions to other certified CDFIs $2,606,310,692 3.6%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
Percentages do not add to 100 because of rounding.
a
Does not include transactions made by banks and credit unions, since these institutions do not report
these data together with other transactions (instead, they report them in aggregate by census tract).
In addition to examining how recipients used funds, we reviewed the
extent of and findings from research on the economic impact of the CDFI
Program and emergency programs. Further, to provide insight on areas
potentially affected by CDFI Program awards, we conducted geographic
and sociodemographic analyses of CDFI Program recipients and project
locations. For the results of this work, see appendix I.
Some Data on Borrower The data on the transactions CDFIs make include several items related to
Characteristics Were the characteristics of individual and business borrowers. These fields are
largely optional, and for some of the items, much of the information is
Unavailable
missing. However, other items are relatively more complete and can
provide some indication of the types of individuals and businesses that
CDFIs lend to.
Individuals. Of the fields related to the characteristics of individual
borrowers, two—income status and credit score—were available for a
majority of transactions to individuals, although the share of unavailable
Page 27 GAO-23-105952 Community Development Financial Institutions
data was substantial for both items. 35 The data for these items suggest
that institutions served borrowers with a wide range of income levels and
access to credit. 36 For income status, of the approximately 70 percent of
transactions for which data were available, nearly half were made to
lower-income individuals, although borrowers at all levels of the income
scale were represented (see table 14).
Table 14: Income Status of Individual Borrowers Served by Institutions Receiving Financial Assistance from the CDFI
Program, Fiscal Years 2018–2020
Percentage of transactions made to individuals in each income category
Fiscal Extremely Very low Low Middle Upper Do not know
year low-income income income income income or missing
2018 5% 8% 21% 23% 13% 30%
2019 4% 7% 26% 15% 16% 32%
2020 3% 6% 20% 15% 31% 25%
2018–2020 4% 7% 22% 17% 22% 28%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
The income categories are defined as follows: extremely low-income means borrower’s family income
is <30 percent of area median family income (AMFI); very low-income means borrower’s family
income is ≥30 percent and <50 percent of AMFI; low-income means borrower’s family income is ≥50
percent and ≤80 percent of AMFI; middle-income means borrower’s family income is >80 percent and
<120 percent of AMFI; upper-income means borrower’s family income is ≥120 percent of AMFI.
35For fiscal years 2016–2020, data on the following borrower characteristics were either
missing or categorized as “do not know” for the majority of loans made to individuals: race,
Hispanic origin, gender, whether the borrower’s household was headed by a single
woman, and whether the borrower had basic banking services at time of intake. Treasury
officials stated that reporting institutions may not collect these data. Regarding race and
ethnicity, the officials said the Equal Credit Opportunity Act limits regulated institutions’
collection of this information for all types of loans except mortgage loans for monitoring
purposes.
36For both characteristics, data are presented for fiscal years 2018–2020 because the
data were either largely missing or used a different reporting scale for fiscal years 2016
and 2017.
Page 28 GAO-23-105952 Community Development Financial Institutions
Similarly, CDFIs reported making loans to individuals with a variety of
credit scores, although the majority went to those with good or very good
credit (see table 15). 37
Table 15: Credit Scores of Individual Borrowers Served by Institutions Receiving Financial Assistance from the CDFI
Program, Fiscal Years 2018–2020
Percentage of transactions made to individuals in each credit score category Do not
Fiscal Excellent credit Very good credit Good credit Fair credit Poor credit No or thin know or
year (800-850) (740-799) (670-739) (580-669) (300-579) credita missing
2018 8% 24% 27% 20% 6% N/A 16%
2019 9% 28% 28% 18% 4% N/A 13%
2020 12% 29% 22% 17% 2% 1% 17%
2018–2020 10% 28% 25% 18% 3% 1% 15%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: The data in the table include institutions that received financial assistance awards from either
the Community Development Financial Institutions (CDFI) Program or the Native American CDFI
Assistance Program. The data do not include consumer loans made by banks or credit unions
because data on such loans are reported separately, aggregated at the census tract level.
A credit score is a numeric value (generally ranging from 300 to 850) that indicates a borrower’s
ability to repay future obligations.
Percentages may not add to 100 because of rounding.
a
This option was introduced in fiscal year 2020 and is defined as borrowers having thin or no credit
profile, such as recent immigrants.
Businesses. Data on entity structure, which were available for most
transactions to businesses from fiscal years 2017 through 2020, indicated
that over 95 percent of these transactions went to for-profit businesses
each year. 38 About 1 to 2 percent of transactions went to nonprofits, while
much smaller numbers went to tribal or village governments—just 65
transactions across the 4-year period, or one-tenth of 1 percent. Another
1 percent of transactions were made to businesses categorized as
“other.” The CDFI Fund also collects data on whether businesses were
37A credit score is a numeric value (generally ranging from 300 to 850) that indicates a
borrower’s ability to repay future obligations. Prior to fiscal year 2020, CDFI Program
recipients reported borrowers’ numeric credit scores, while in fiscal year 2020, they
reported the category in which the borrower’s credit score fell (e.g., excellent credit, very
good credit, etc.). For our analysis, we collapsed the older numeric scores into the
categories.
38Data on entity structure were missing for 69 percent of transactions to businesses in
fiscal year 2016, but the share of missing data decreased to 14 percent in fiscal year 2017
and then 0 percent for fiscal years 2018 through 2020, as the CDFI Fund began requiring
CDFIs to report these data for business borrowers.
Page 29 GAO-23-105952 Community Development Financial Institutions
more than 50 percent owned or controlled by minorities, women, and low-
income individuals, but missing data and changes over time in the
instructions for reporting this information limits their use for the fiscal year
2016–2020 period. 39
Treasury Responded
to Potential Risks
When Creating
Emergency Programs
Treasury Took Steps to Treasury based the two emergency grant programs–RRP and ERP–on its
Address Potential Risks existing financial assistance award process for the CDFI Program. For
example, applicants for these three programs used the same automated
Introduced by Expanded
system to submit application material and underwent similar financial
Eligibility and Other analyses and compliance risk evaluations within the automated system. 40
Features of Emergency Recipients for each of the programs have or will be required to enter into
Grant Programs a legally binding assistance agreement that outlines the performance
goals and reporting requirements. In addition, all of the programs
generally will have the same post-award reporting requirements for the
performance period of the awards.
However, both emergency grant programs differed from the CDFI
Program financial assistance awards process in ways that may introduce
risks to program integrity, as follows.
Eligibility of riskier institutions. Relative to the CDFI Program, RRP
and ERP allowed institutions of higher financial risk to be considered for
awards. Treasury officials told us they expanded eligibility to be more
inclusive of institutions that may have received less favorable risk scores
because of the risk these institutions incur to better address the needs of
the low-income communities they serve. For example, Treasury officials
39If information for these three items was not known or not applicable, the guidance either
instructed CDFIs to leave the items blank, select “No,” or select “Do Not Know,”
depending on the year (the “Do Not Know” option was introduced in fiscal year 2020).
These differences across years make interpretation of the data difficult. Further, for some
years, much of the information for these items was missing or was recorded as “Do Not
Know.”
40Although Treasury has not yet awarded ERP funds or finalized procedures for the
program, it has published program requirements governing the award process in the
Federal Register.
Page 30 GAO-23-105952 Community Development Financial Institutions
said some MDIs have a less favorable risk score because of low levels of
deposits or high levels of residential mortgages, which can result in lower
than average performance ratios. According to Treasury, 57 of the 863
RRP recipients received a less favorable risk score from either their
regulator or from the CDFI Fund, as explained below.
Three scoring systems measuring financial risk have implications for
awards. CAMELS is an acronym that refers to a scoring system used by
bank supervisors to rate institutions on capital adequacy, asset quality,
management, earnings, liquidity, and sensitivity to market risk. The
National Credit Union Administration has used a similar scoring system,
CAMEL, that generally rates institutions on the same components, with
the exception of sensitivity to market risk. 41 The CDFI Fund has its own
scoring system to measure financial risk for institutions that do not receive
a CAMELS or CAMEL score–the CDFI Fund financial composite score.
The score is based on an analysis of 23 financial indicators. All three
scoring systems have a range of 1 through 5, with higher numbers
indicating weaker financial condition and risk management.
Under RRP and ERP, regulated institutions with CAMELS or CAMEL
scores of up to 4 and unregulated institutions with CDFI Fund financial
composite scores of up to 4 are eligible for awards. In contrast, only
institutions with scores of 1 through 3 are eligible for CDFI Program
financial assistance awards. 42
Applicants in the higher-risk category went through an additional step in
the application review and award determination process to address the
potential risk. Treasury procedures required all applicants that passed the
financial analysis and compliance risk evaluation in the RRP application
review process to undergo a due diligence review. Only applicants
considered higher risk received an additional written due diligence
analysis by multiple staff, according to Treasury procedures and program
staff. These included applicants with a CAMELS, CAMEL, or CDFI Fund
financial composite score of 4, audit findings, or other adverse indicators.
The RRP due diligence review procedures considered the applicant’s
financial stability, history of performance in managing federal awards,
41Beginning April 1, 2022, the National Credit Union Administration added sensitivity to
market risk as a component of its rating.
42Institutions with a CAMELS or CAMEL score of 4 are eligible for CDFI Program technical
assistance awards.
Page 31 GAO-23-105952 Community Development Financial Institutions
audit or regulator findings, and the applicant’s ability to effectively
implement federal requirements. According to Treasury procedures,
applicants that received a “very concerned” rating on these reviews were
unqualified for awards, and applicants that received a “somewhat
concerned” rating were to have their award amount decreased by 30
percent. ERP requirements also allow for Treasury reviewing officials to
reduce the award amount for applicants with these scores.
Increased award size. RRP and ERP generally offer larger awards than
the CDFI Program. The median RRP award (which, as previously
discussed, equaled the maximum award) was about $1.8 million, while
the median CDFI Program award for fiscal year 2021 was $400,000. 43
Although ERP award amounts have not yet been determined, the
maximum award size is $15 million and the minimum award size is
$500,000, for use over a 5-year period. 44 Further, 292 RRP recipients
also received a fiscal year 2021 CDFI Program award, and 27 institutions
received a fiscal year 2021 CDFI Program award, an RRP award, and an
ECIP award. Some of these institutions may also be eligible for an ERP
award. Institutions that received multiple awards had significantly more
award funds to use and manage than they would have had under the
CDFI Program alone.
Treasury addressed potential capacity risks by imposing caps on award
size that scale to the recent business volume of the institution. For RRP,
award amounts could not exceed the greater of either 1.5 times the
applicant’s total on-balance-sheet financial products closed in eligible or
target markets in its most recent fiscal year, or the minimum award
amount of $200,000. 45 Treasury officials said they thought this amount
was a good proxy to represent the potential amount of business an
institution would have the capacity to do during the 2-year performance
period. Officials said they based this on their knowledge of historic
patterns of award fund use by CDFIs. To build capacity, minimum awards
43The median fiscal year 2021 CDFI Program award was smaller than prior years’ awards.
For fiscal years 2019–2021, the highest median award amount was $565,000.
44Depending on the number of qualified applicants, the maximum award amount may be
significantly less than $15 million.
45An eligible market is (i) a geographic area meeting the requirements set forth in 12 CFR
1805.201(b)(3)(ii), or (ii) individuals who are low-income, African American, Hispanic,
Native American, Native Hawaiians residing in Hawaii, Alaska Natives residing in Alaska,
or Other Pacific Islanders residing in American Samoa, Guam or the Northern Mariana
Islands. Target markets are the markets that a CDFI is approved to serve when receiving
CDFI certification.
Page 32 GAO-23-105952 Community Development Financial Institutions
of $200,000 could be spent entirely on eligible operational activities. For
ERP, award amounts may not exceed 3 times the applicant’s total on-
balance-sheet financial products closed in its 5 most recent fiscal years.
Treasury officials said the cap was generally modeled on the cap for
RRP, adjusted for the longer 5-year performance period of ERP.
Additionally, Treasury officials said they will consider the administration of
prior awards in assessing an institution’s capacity to use an ERP award.
According to program requirements, ERP applicants will receive a score
for “organizational capacity,” which includes factors such as whether the
current and proposed staffing is sufficient to support the proposed
business strategy. Application reviews will also consider performance risk
with respect to all previous CDFI Fund awards. Finally, reviewing officials
may lower the recommended award amount if there are concerns about
the applicant’s ability to deploy the full award amount.
Additional targeting requirements for ERP. The authorizing statute for
ERP states that the funds are meant “to expand lending, or grant making
in low- or moderate-income minority communities and to minorities that
have significant unmet capital or financial service needs.” 46 Further, the
statute states that all three of the emergency programs are meant to
serve communities that have disproportionately suffered from the impacts
of the COVID-19 pandemic. 47 These specific targeting requirements
potentially increased Treasury’s compliance risk compared to the regular
CDFI program, which does not have such requirements.
To help ensure funds reach the communities intended by statute,
Treasury added certain eligibility requirements and steps in the
application review and award processes. Treasury limited eligibility to: (1)
institutions serving census tracts that it determined had been severely
impacted by COVID-19, had a median income at or below 120 percent of
the area median income, and were a CDFI Investment Area; and (2)
46Pub. L. No. 116-260, div. N, § 522(a)(2), 134 Stat. 1182, 2088 (2020).
47Pub. L. No. 116-260, div. N, § 520, 134 Stat. 1182, 2079 (2020).
Page 33 GAO-23-105952 Community Development Financial Institutions
institutions serving Native Areas. 48 To be eligible, applicants must
demonstrate that at least 30 percent of their average annual financial
products closed or grants made over their 5 most recent fiscal years went
to ERP-eligible geographies. When reviewing applications, an external
reviewer must consider criteria, such as
• whether the applicant demonstrates a strong understanding of the
economic impact of the COVID-19 pandemic on the geographies and
populations it is proposing to serve;
• whether the applicant has engaged with and obtained input on its
business strategy from these populations; and
• whether the applicant proposed to use culturally and linguistically
relevant marketing and outreach efforts that will support the
deployment of the award consistent with its business strategy.
Further, recipients must sign assistance agreements that specify that at
least 90 percent of ERP award funds will be used in the designated
eligible geographies and that any remaining funds will serve low- and
moderate-income people or businesses disproportionately impacted by
the COVID-19 pandemic. In addition, the authorizing statute states that
participating institutions may request data from borrowers and applicants
for credit that will allow them to ensure the targeted populations and low-
income residents are adequately served. Treasury officials said they will
require institutions to collect race and ethnicity data for all borrowers or
other beneficiaries of financial products and services supported by ERP.
Streamlined application review procedures for RRP. To meet RRP’s
expedited award time frames required by the Consolidated Appropriations
Act, 2021, Treasury streamlined its existing process for evaluating CDFI
Program financial assistance applications. Specifically, Treasury removed
the business plan review and policy objective review steps used in
evaluating financial assistance applications for the CDFI Program
because the act did not require them. According to Treasury documents,
48Treasury considered a census tract to have experienced ‘‘severe impact’’ of the COVID-
19 pandemic if it met one or more of the following criteria: (a) demonstrated severe
mortality, based on being in the highest tercile of the number of deaths per 100,000
people, according to reported cumulative mortality for the period from April 1, 2020, to
March 31, 2021, based on data from the Centers for Disease Control and Prevention and
the Government of Puerto Rico; (b) demonstrated severe change in unemployment; or (c)
demonstrated low community resilience, based on being in the highest tercile of the
percentage of individuals or families that have three or more resilience-related risk factors
relative to the impact of disasters such as pandemics, based on data from the U.S.
Census Bureau, Community Resilience Estimates Program.
Page 34 GAO-23-105952 Community Development Financial Institutions
the intent of the business plan review is to ensure that each applicant’s
business plan is sound and achievable. The policy objective review
evaluates the ability of the applicant to meet the policy objectives of the
program.
While RRP applicants did not receive a business plan or policy objective
review, the evaluation process and program requirements included
several measures to help ensure the institutions were financially stable
and able to use the awards as intended. In addition to the financial
analysis and due diligence reviews discussed above, program guidance
called for each applicant to receive an eligibility review and a compliance
risk evaluation. The compliance risk evaluation was meant to use a
variety of information to help ensure the institution had no major internal
management or compliance concerns. In addition, Treasury officials said
they added a requirement that regulated entities have a Community
Reinvestment Act rating of “satisfactory” or better to be eligible for the
program because of the streamlined review process. 49 Further, Treasury
officials said that they limited the award amount for applicants without
audited financial statements to $200,000 to help limit risk. According to
Treasury officials, the business plan and policy objective reviews were not
needed for RRP because the program was designed to provide funding to
all eligible applicants based on a formula that considers specific criteria
and the applicant’s historical business volume, whereas CDFI Program
financial assistance applications are competitive.
Treasury has begun to conduct post-award compliance monitoring for
RRP, which will provide information on whether recipients are managing
funds effectively and using funds for intended purposes. Similar to the
CDFI Program, Treasury requires RRP recipients to submit annual
financial condition and performance reports to help monitor compliance
with program requirements and goals. Performance goals include closing
financial products equal to at least 50 percent of the award amount in
eligible or target markets by the end of the first full fiscal year after the
award was made, and at least 70 percent of the award amount in eligible
or target markets by the end of the second year of the performance
period.
49The Community Reinvestment Act, 12 U.S.C. §§ 2901-2908, encourages depository
institutions to meet the credit needs of communities where they operate. Federal financial
regulators are required to assess periodically and rate each bank’s record of helping to
meet the credit needs of its entire community.
Page 35 GAO-23-105952 Community Development Financial Institutions
Treasury Considered The Emergency Capital Investment Program—through which Treasury
Certain Risks in Designing made capital investments in financial institutions in exchange for
preferred stock or subordinated debt from the institutions—is
the Emergency Capital
substantively different from the CDFI Program and the emergency grant
Investment Program programs. For example, ECIP has a narrower universe of eligible
institutions—applicants must be federally insured depository institutions,
bank or savings and loan holding companies, or federally insured credit
unions and either a CDFI or MDI. In addition, institutions that receive an
investment must pay Treasury dividends or interest after the first 2 years
for the life of the investment. A capital investment also benefits an
institution differently than a grant, according to Treasury officials.
Specifically, they told us a capital investment may support long-term
growth by allowing the institution to take on additional funding and
increase its balance sheet.
Program integrity risks for ECIP also varied somewhat from those for the
emergency grant programs. Treasury took steps to address these risks
when developing its application review and reward determination
processes, as follows.
Ability to pay dividends or interest. Institutions that received capital
investments entered into a legal agreement with Treasury to pay either
dividends or interest, depending on the type of institution, for the life of
the investment. 50 Treasury took several steps during the application
review process to help ensure any institution offered an investment would
be able to pay the required dividends or interest. As a measure of the
institution’s ability to pay the required annual dividend or interest amount,
application reviewers were required to calculate a ratio of the institution’s
projected annual net income (using the most recent 5-year average return
on average assets) to the annual dividend or interest amount. Treasury
required this ratio to be at least 1 to 1. According to program guidance, an
external financial agent completed this assessment and Treasury
analysts reviewed it. If an institution did not meet the required ratio,
Treasury could lower its investment amount to satisfy the requirement.
According to Treasury, the underwriting process also assessed the
institutions’ ability to execute its lending plan, which included information
such as the institution’s business strategy and operating goals.
Institutional capacity. Institutions were expected to increase lending
with the capital investment Treasury provided, in some cases
50Dividends and interest are paid quarterly after the first 2 years of the investment.
Page 36 GAO-23-105952 Community Development Financial Institutions
substantially. The application included a section on the institution’s growth
strategy, in which the institution specified the dollar amount of increased
lending to targeted groups that it would conduct over 10 years as a direct
result of the capital investment it sought from Treasury. In addition, ECIP
investments were sometimes large, up to $250 million. Similar to the
emergency grant programs, Treasury needed assurance that institutions
had the capacity to increase lending and use the large awards effectively.
Treasury procedures called for review of several measures of institutional
capacity to ensure that ECIP applicants would be able to execute their
lending plans submitted as part of the application. The lending plans
included information such as the projected loans to be originated in the
next 3 years to certain targeted populations, the institution’s business
strategy and operating goals, and its growth strategy. Applicants received
scores on four measures related to capacity to execute on the plan. One
score specifically measured the “internal capacity” of the applicant,
including staffing levels. The other scores related to past experience in
managing federal awards, the reasonableness of lending plan projections
based on market size and condition, and the applicant’s track record of
lending to the targeted populations based on loans made over the past 2
fiscal years. In addition, Treasury officials said they provided an
opportunity for the federal banking regulators and the National Credit
Union Administration to raise any concerns about Treasury investing the
full amount requested by the applicant.
Other program risks. Treasury identified and responded to two other
potential program risks: (1) the risk that institutions would use award
funds to compensate shareholders or officers instead of increasing
lending and (2) the risk that recipients would fail to maintain certification
as a CDFI or MDI or fail to submit timely and accurate reports.
According to Treasury officials, Treasury has addressed the first risk
through an interim final rule restricting the use of funds for executive
compensation, share buybacks, and dividends. 51 In the overview of the
interim final rule, Treasury states that the restrictions seek to promote the
integrity of ECIP by ensuring that institutions use the award funds to
provide loans, grants, and forbearance for small businesses, minority-
owned businesses, and consumers, especially in low-income and
underserved communities. Treasury officials said the second risk is
51The Consolidated Appropriations Act, 2021 required Treasury to issue rules setting
restrictions on executive compensation, share buybacks, and dividend payments. 86 Fed.
Reg. 13449 (Mar. 9, 2021) (codified at 31 C.F.R. pt. 35)
Page 37 GAO-23-105952 Community Development Financial Institutions
addressed by ECIP legal agreements that institutions must enter into as a
condition of receiving funding. For example, the securities purchase
agreement for senior preferred stock specifies that when providing
ongoing supplemental reports to Treasury, the principal executive officer
or principal financial officer of the recipient must certify that the
information in the report is accurate. In addition, the recipient must
annually certify that the processes and controls used to generate the
report are satisfactory.
According to program officials, Treasury is finalizing ongoing reporting
requirements and policies and procedures that it will use to conduct post-
award monitoring, which will be necessary to help ensure the integrity of
the program.
Treasury Assessed
Improper Payment Risk
and Is Developing
Additional Processes to
Address Fraud Risk
Improper Payment Risk Treasury conducted an improper payment risk assessment for RRP in
Assessments fiscal year 2022, as required by the Payment Integrity Information Act of
2019, and determined the program was not susceptible to significant
improper payments. 52 Treasury has not yet completed an assessment for
ERP or ECIP. Under the Office of Management and Budget’s current
improper payment estimate guidance, agencies are not required to
estimate improper payments for programs in their initial year of operation.
Specifically, according to the guidance, agencies should complete a risk
assessment to determine susceptibility to significant improper payments
52Improper payments are payments that under statutory, contractual, administrative, or
other legally applicable requirements should not have been made or were made in an
incorrect amount. The purpose of an improper payment risk assessment is to determine
whether a program is susceptible to significant improper payments. A determination of
susceptibility triggers a requirement to report an improper payment estimate in the
following fiscal year. 31 U.S.C. § 3352(c)(1). The Payment Integrity Information Act of
2019 defines significant improper payments as improper payments in the preceding fiscal
year that may have exceeded either (1) 1.5 percent of program outlays and $10 million or
(2) $100 million (regardless of the improper payment rate).
Page 38 GAO-23-105952 Community Development Financial Institutions
after the first 12 months of program operations. 53 ERP and ECIP have not
yet reached this threshold. Treasury officials said they have not yet
completed an assessment for ERP and ECIP because one is not required
until fiscal year 2023.
Fraud Prevention Measures Treasury has some fraud prevention measures in place for the
emergency programs, and it is developing a new process that it will use
for all CDFI Fund programs, including the emergency grant programs.
RRP and ERP. Treasury officials said they have fraud prevention
measures in place for RRP and ERP both before and after disbursing
funds. For example, before disbursement, Treasury requires institutions
to enter into legally binding assistance agreements. These agreements
lay out the award terms and conditions, post-award reporting
requirements, and specific performance goals for the institution. By
signing the agreement, an authorized representative attests to certain
facts about the institution and agrees to administer the award in
accordance with the terms and conditions, which include a requirement to
report any suspected fraud, waste, or abuse of award funds to Treasury’s
Office of Inspector General (OIG).
After disbursement, Treasury may conduct ad hoc desk reviews and site
visits of institutions that have received CDFI Program and RRP awards.
The desk reviews and site visits are designed to detect mismanagement
and abuse, in addition to reviewing compliance with award requirements
in assistance agreements and providing technical assistance, according
to Treasury’s standard operating procedures. In addition, Treasury
officials said they conduct an internal controls assessment every year as
part of the agency’s annual financial statement audit, which assesses
whether proper controls are in place to prevent fraud, waste, and abuse in
each of Treasury’s programs.
ECIP. Treasury officials told us they implemented mechanisms to prevent
fraud in ECIP during the application review process and when finalizing
agreements and disbursing funds to institutions. Treasury’s application
review procedures called for the use of various mechanisms to ensure
53Office of Management and Budget, OMB Circular A-123 Appendix C: Requirements for
Payment Integrity Improvement, M-21-19 (Washington, D.C.: Mar. 5, 2021). In prior work,
we recommended that Congress amend the Payment Integrity Information Act, 2019 to
designate all new programs making more than $100 million in payments in a fiscal year
automatically susceptible to significant improper payments. See GAO, Emergency Relief
Funds: Significant Improvements Are Needed to Ensure Transparency and Accountability
for COVID-19 and Beyond, GAO-22-105715 (Washington, D.C.: Mar. 17, 2022).
Page 39 GAO-23-105952 Community Development Financial Institutions
institutions were eligible for federal funding, officials said. For example,
officials told us that they used ID.me (an online identity authentication
service) to authenticate the identities of the applicants. They also
consulted with federal banking regulators and the National Credit Union
Administration about institutions’ eligibility to participate. Before
disbursing funds, Treasury’s procedures called for staff to take several
steps to help reduce the risk of improper payments. These included
reviewing updated information about the institutions to confirm they were
still eligible to participate in the program and checking to ensure the final
funding amount was less than or equal to the preliminary approval
amount. According to program officials, another factor helping prevent
waste in the program is the interim final rule discussed earlier, which
restricts the use of award funds for executive compensation, share
buybacks, and dividends.
RRP and ERP post-award monitoring. Treasury is in the process of
developing a risk-based approach for conducting desk reviews and site
visits that it will use for the CDFI Program, RRP, and ERP. For example,
Treasury hired a contractor to develop a suite of tools to identify high-risk
institutions, allowing Treasury to prioritize those institutions for desk
reviews and site visits. The suite of tools includes the application
assessment tool, which is already in use, and a noncompliance
scorecard. The noncompliance scorecard will identify award recipients at
risk of noncompliance or default based on their track record for meeting
reporting requirements, performance goals and measures, and
noncompliance with any other portions of their assistance agreements.
Treasury officials said they had not yet finalized the specific variables for
the scorecard. The officials also said they were finalizing the policies and
procedures for the new process, but they did not have an estimated
implementation date. These efforts are intended to monitor whether the
funds are being used as intended.
As of October 2022, Treasury OIG had completed two evaluations of the
emergency programs and had additional reviews ongoing.
• In December 2021, OIG issued a report on Treasury’s design and
implementation of RRP, including policies and procedures for
selecting recipients. The audit found that Treasury established internal
controls for award determination in accordance with Standards for
Internal Controls in the Federal Government, but also that Treasury’s
assistance agreement template was missing certain terms and
Page 40 GAO-23-105952 Community Development Financial Institutions
conditions regarding recipient integrity and performance matters. 54
The report noted that Treasury subsequently revised the template
before issuing RRP awards.
• In March 2022, Treasury OIG issued a report on Treasury’s
implementation of ECIP, including Treasury’s efforts to develop
eligibility and application requirements and policies and procedures
for the program prior to disbursement. 55 The review found that
Treasury worked swiftly to implement ECIP, although it did not meet
the statutory deadline for accepting applications and had not
completed policies and procedures for post-investment monitoring.
OIG recommended that Treasury finalize policies and procedures and
key documentation to govern full program implementation and
ongoing administration of ECIP investments, including internal
controls over the program. Treasury took several steps to address this
recommendation. In April and May 2022, it finalized policies and
procedures for application intake and review, award determinations,
closing the investments, and disbursing funds. As of September 2022,
Treasury was also finalizing procedures for quarterly supplemental
reporting by recipients, monitoring recipients’ compliance with
investment terms, setting dividends and interest rates, and processing
payments from institutions, according to program officials.
Treasury OIG is conducting two additional audits of the emergency grants
programs. In May 2022, OIG began an audit of Treasury’s implementation
of ERP. The review will look at steps Treasury took to develop policies
and procedures and make award funds available, among other things. In
June 2022, OIG began an audit of Treasury’s award and post-award
administration of RRP, including processes for ensuring the accuracy of
RRP payment and post-award monitoring of recipients.
We provided a draft of this report to the Department of the Treasury for
Agency Comments review and comment. Treasury provided technical comments, which we
incorporated as appropriate.
54Department of the Treasury, Office of the Inspector General, Coronavirus Disease 2019
Pandemic Relief Programs: Audit of the Community Financial Development Fund’s
Implementation of CDFI Rapid Response Program, OIG-22-023 (Washington, D.C.: Dec.
21, 2021).
55Department of the Treasury, Office of the Inspector General, Coronavirus Disease 2019
Pandemic Relief Programs: Audit of Treasury’s Implementation of the Emergency Capital
Investment Program, OIG-22-028 (Washington, D.C.: Mar. 8, 2022).
Page 41 GAO-23-105952 Community Development Financial Institutions
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 to the appropriate
congressional committees and the Secretary of the Treasury. In addition,
the report will be available at no charge on the GAO website at
https://www.gao.gov.
If you or your staff have any questions about this report, please contact
me at (202) 512-8678 or shearw@gao.gov. Contact points for our Offices
of Congressional Relations and Public Affairs may be found on the last
page of this report. GAO staff who made key contributions to this report
are listed in appendix IV.
Sincerely Yours,
William B. Shear
Director, Financial Markets and Community Investment
Page 42 GAO-23-105952 Community Development Financial Institutions
Appendix I: Available Research on Economic
Appendix I: Available Research on Economic
Impact of Programs and Geographic Analysis
Impact of Programs and Geographic
of CDFI Program Recipients
Analysis of CDFI Program Recipients
Availability of Economic Research on the economic impact of the Community Development
Impact Research Financial Institutions (CDFI) Program is limited. Estimating the economic
impact of government programs (for example, changes in employment
rates or income levels caused by the program) is challenging because it
involves comparing program outcomes to outcomes that would have
occurred in the program’s absence. Making causal inferences about
program impacts typically requires rigorous comparison of treatment and
control groups, but this can be difficult to achieve because of data
limitations and other factors. 1
Through a literature search, we identified just one study, published in
2013, that provided information about the impact of the CDFI Program. 2
This study analyzed the effects of receiving an award from the program
on CDFIs’ lending activity and growth. The sample was limited to credit
unions that applied for CDFI Program funding from 2000 through 2009, so
it is not generalizable to all program recipients. Findings from the study
included the following:
• Receiving an award from the program catalyzed additional lending by
credit unions. Specifically, the authors estimated that for each dollar
received from the program, credit unions lent out an additional $0.45
in the first year after receiving the award, $1.10 after 2 years, and
$1.60 after 3 years.
• Receiving an award also led to increases in deposit rates and net
worth growth among credit unions. For each dollar received from the
program, $0.17 went toward growth in recipients’ net worth.
1Random assignment to either the program or a control group minimizes any preexisting
differences between the groups, helping ensure that differences in outcomes are
attributable to the program and not other factors. However, using random assignment to
determine program participation is often impractical. An alternative approach is a quasi-
experimental design in which evaluators compare participants to another group that is
thought to be similar in key ways. A limitation of quasi-experiments is the potential for
preexisting differences between groups, which make it difficult to disentangle the effects of
the program from other factors. Evaluators can use statistical methods to attempt to
control for such differences, but it is not possible to control for all potential differences.
2Kristle Romero Cortés and Josh Lerner, “Bridging the Gap? Government Subsidized
Lending and Access to Capital,” Review of Corporate Finance Studies, vol. 2, no. 1
(2013): 98–128. https://doi.org/10.1093/rcfs/cft002. Our literature review identified two
additional studies that sought to evaluate the impact of the CDFI Program, but both had
methodological limitations that prevented us from drawing conclusions about the
program’s impact. Appendix II provides more information about how we conducted our
literature review.
Page 43 GAO-23-105952 Community Development Financial Institutions
Appendix I: Available Research on Economic
Impact of Programs and Geographic Analysis
of CDFI Program Recipients
• Despite these positive findings, the amount of delinquent loans
recipients made increased. For each dollar received from the
program, delinquent loan growth increased by $0.12 over the 3-year
period following the award.
No research has evaluated the impacts of three emergency programs
developed to help CDFIs respond to the impacts of COVID-19—the Rapid
Response Program (RRP), Equitable Recovery Program (ERP), and
Emergency Capital Investment Program (ECIP). At the time we
conducted our audit work, these programs had not been in existence long
enough to measure their effects. The Department of the Treasury made
RRP awards in June 2021 with a 2-year period of performance, and it
planned to announce ERP awards in the winter of 2023 (following an
application deadline of September 22, 2022). In September 2022,
Treasury announced it had completed a substantial amount of the ECIP
investments.
The Consolidated Appropriations Act, 2021, required Treasury to conduct
a study of the emergency programs’ impact on low- and moderate-income
and minority communities and submit a report to Congress on the results
of the study by June 27, 2022. However, the report Treasury submitted
noted that the emergency programs were still early in their
implementation, so the data needed to meaningfully report on their
impacts were not yet available. Instead, the report described the
implementation of the three programs and provided data on the
institutions selected to receive funding from two of the programs, RRP
and ECIP. For example, it reported information on institution type and the
distribution of awards or investment offers by state. The report also stated
that Treasury planned to gather data from award recipients consistent
with statutory requirements.
As of mid-2022, CDFI Fund and ECIP officials stated they had not yet
finalized plans to evaluate the emergency programs. However, CDFI
Fund officials said that further analyses of RRP will be conducted after
recipients submit data on outcomes and on the transactions they make
during the program’s reporting period. The officials had not yet
determined who would conduct these additional analyses. Further, ECIP
officials stated they were in the process of determining an evaluation
approach for the program, which they said would likely focus on the
extent to which institutions that receive investments from the program
increase their lending above baseline levels. This kind of focus could help
ECIP officials determine whether the program stimulates additional
private investment.
Page 44 GAO-23-105952 Community Development Financial Institutions
Appendix I: Available Research on Economic
Impact of Programs and Geographic Analysis
of CDFI Program Recipients
Characteristics of To provide insight on locations potentially affected by CDFI Program
Communities Served by awards, we analyzed geographic, demographic, and socioeconomic
characteristics of communities served by award recipients ranging from
CDFI Program Recipients
fiscal years 2016 through 2021. We analyzed the metropolitan and
nonmetropolitan breakdown of fiscal year 2019–2021 recipient locations,
the demographic and socioeconomic characteristics of the counties in
which they were located, and the dollar amount of awards by state. 3 We
also analyzed the demographic and socioeconomic characteristics of the
census tracts where CDFIs had ongoing projects.
From fiscal years 2019 through 2021, about 80 percent of CDFI Program
recipients were located in metropolitan areas and the remainder were
located in nonmetropolitan areas (see table 16).
Table 16: Location Type of Community Development Financial Institutions (CDFI) Program Recipients, Fiscal Years 2019–2021
CDFIs located in CDFIs located in
metropolitan areas nonmetropolitan areas Total
Number 398 107 505
Percentage 78.8% 21.2% 100%
Source: GAO analysis of Department of the Treasury and U.S. Department of Agriculture data. | GAO-23-105952
Counties where CDFI Program recipients were located were generally
more diverse than the counties where there were no recipients (see table
17). In particular, counties where CDFIs were located had, on average,
higher percentages of non-White, Hispanic or Latino, and foreign-born
residents. In addition, the unemployment and poverty rates in those
counties were higher than in the counties without recipients. However,
median household incomes were higher in counties where recipients
were located and a higher percentage of households had internet access.
3Institutions only reported one location in their annual certification reports. For institutions
with multiple branches, the data do not include additional locations beyond the one
reported in the certification reports. CDFIs may make investments in communities outside
of their reported location.
Page 45 GAO-23-105952 Community Development Financial Institutions
Appendix I: Available Research on Economic
Impact of Programs and Geographic Analysis
of CDFI Program Recipients
Table 17: Average Socioeconomic and Demographic Characteristics of Counties in Which Community Development Financial
Institutions (CDFI) Program Recipients Were and Were Not Located, Fiscal Years 2019–2021
Counties where CDFIs Counties where CDFIs
were located were not located
Estimated mean Estimated mean
Number of (confidence Number of (confidence
County characteristic counties interval) counties interval)
Percentage of residents who were non-White 293 31.4 2,850 16.9
(29.1–33.7) (16.3–17.5)
Percentage of residents who were Hispanic or Latino 293 12.0 2,850 9.3
(10.2–13.8) (8.8–9.8)
Percentage of residents who were foreign-born 293 8.9 2,850 4.3
(7.8–9.9) (4.1–4.5)
Unemployment rate 293 5.9 2,850 5.1
(5.6–6.2) (5.0–5.2)
Poverty rate 293 15.2 2,850 14.4
(14.5–15.9) (14.2–14.6)
Median household income 293 $59,898 2,849 $54,526
($57,704–$62,092) ($54,010–$55,042)
Percentage of households with no internet access 293 15.2 2,850 18.2
(14.3–16.1) (18.0–18.5)
Source: GAO analysis of Department of the Treasury and Bureau of the Census data. | GAO-23-105952
Notes: Ninety-five percent confidence intervals are presented in parentheses.
County characteristics reflect American Community Survey estimates for 2016–2020.
The amount CDFI Program recipients received from fiscal years 2016
through 2021 varied by state (see fig. 5). Over that period, ten states—
California, Florida, Illinois, Louisiana, Massachusetts, Minnesota,
Mississippi, New York, Ohio, and Pennsylvania—were in the top quintile
for award amounts, each receiving more than $34 million.
Page 46 GAO-23-105952 Community Development Financial Institutions
Appendix I: Available Research on Economic
Impact of Programs and Geographic Analysis
of CDFI Program Recipients
Figure 5: Distribution of Community Development Financial Institution Program Awards by State, Fiscal Years 2016–2021
Census tracts where CDFI Program recipients had active projects were
generally more diverse than census tracts where CDFIs were not
conducting projects (see table 18). On average, tracts where recipients
had active projects had a higher percentage of residents that were non-
White, Hispanic or Latino, or foreign-born. However, the median housing
value was higher in census tracts where recipients had active projects.
Page 47 GAO-23-105952 Community Development Financial Institutions
Appendix I: Available Research on Economic
Impact of Programs and Geographic Analysis
of CDFI Program Recipients
Median household incomes were similar in census tracts that did and did
not have projects.
Table 18: Average Socioeconomic and Demographic Characteristics of Census Tracts Where Community Development
Financial Institutions (CDFI) Did or Did Not Conduct Projects, Fiscal Years 2016–2020
Census tracts where CDFI Census tracts where CDFIs
projects were active were not conducting projects
Number of Estimated mean Number of Estimated mean
census (confidence census (confidence
Census tract characteristic tracts interval) tracts interval)
Percentage of residents who were non-White 52,690 30.4 20,608 21.8
(30.2–30.6) (21.5–22.1)
Percentage of residents who were Hispanic or 52,690 19.5 20,608 13.0
Latino (19.3–19.7) (12.7–13.3)
Percentage of residents who were foreign-born 52,308 14.3 20,102 8.4
(14.2–14.4) (8.3–8.6)
Unemployment rate 52,671 5.9 20,523 5.9
(5.9–6.0) (5.8–5.9)
Poverty rate 52,665 15.1 20,484 14.9
(15.0–15.2) (14.7–15.0)
Percentage of residents on public assistance 52,656 2.7 20,457 2.5
(2.7–2.7) (2.4–2.5)
Median household income 52,588 $66,590 20,325 $66,373
($66,309–$66,872) ($65,885–$66,861)
Median housing value 51,954 $286,789 20,004 $230,889
($284,630– ($227,737–
$288,948) $234,041)
Source: GAO analysis of Treasury and Bureau of the Census data. | GAO-23-105952
Notes: Ninety-five percent confidence intervals are presented in parentheses.
Census tract characteristics reflect American Community Survey estimates for 2015–2019.
Page 48 GAO-23-105952 Community Development Financial Institutions
Appendix II: Objectives, Scope, and
Appendix II: Objectives, Scope, and
Methodology
Methodology
Our objectives were to examine (1) the types of institutions that received
funds from the Community Development Financial Institution (CDFI)
Program or the 2021 emergency programs and how much they received,
(2) how recipients have used CDFI Program and Rapid Response
Program (RRP) funds, and (3) steps the Department of the Treasury took
to address risks posed by the emergency programs.
Analysis of Institution For the first two objectives, we analyzed several types of Treasury data.
Types, Award Amounts, To determine the types of institutions that received funds from the CDFI
Program or the 2021 emergency programs and how much they received,
and Uses of Funds
we obtained data on award recipients and amounts from the CDFI Fund’s
Awards Management Information System (AMIS) and the Emergency
Capital Investment Program’s (ECIP) data systems.
The AMIS data included recipient and award information for the CDFI
Program for fiscal years 2019–2021 (2021 was the most recent fiscal year
for which CDFI Program data were available), similar information for the
Rapid Response Program (RRP) for fiscal year 2021 (the only year in
which the program received funds), and CDFI-reported Annual
Certification and Data Collection Report (ACR) data for fiscal years 2019–
2020 that included characteristics of recipient institutions such as type
and asset size. We started our CDFI Program analysis with fiscal year
2019 data because earlier ACR data were unreliable for the analyses we
performed. For ECIP, we obtained data on recipient organization type,
asset size, and award amount for funds appropriated in fiscal year 2021
(the only year in which the program received funds).
We analyzed the fiscal year 2019–2021 CDFI Program data to determine:
• the types and number of CDFIs, including Native CDFIs, that received
awards;
• the minimum and maximum values of recipients’ assets, revenue, and
full-time equivalent staff;
• the distributions (including minimum, median, and maximum values)
for award amounts; and
• the primary lines of business and types of development and financial
services recipients offered.
Our CDFI Program analyses included data on institutions that received
financial or technical assistance awards from the CDFI Program or the
Native American CDFI Assistance Program, as well as those that
Page 49 GAO-23-105952 Community Development Financial Institutions
Appendix II: Objectives, Scope, and
Methodology
received supplemental assistance awards for the Healthy Food Financing
Initiative or Persistent Poverty Counties.
For the analyses involving characteristics of award recipients, we
matched CDFI Program awards data for fiscal years 2019–2021 to ACR
data for fiscal years 2019–2020. Our analyses reflect the approximately
68 percent of recipients we were able to match to ACR data. We could
not match all recipients, in part because (1) fiscal year 2021 ACR data
were not yet available when we conducted our analyses and (2) some
institutions that received awards in fiscal years 2019–2021 were not yet
certified and therefore not required to submit ACR data. 1 We used the
matching process to determine recipient characteristics such as
organization type, assets, revenue, full-time equivalent staff, primary lines
of business, and development and financial services offered.
For the emergency programs (RRP and ECIP), we analyzed program
data to determine:
• the types and number of CDFIs that received RRP awards,
• the types and number of CDFIs and minority depository institutions
that received ECIP awards, and
• the distributions (including minimum, median, and maximum values)
for ECIP recipients’ assets and RRP and ECIP award sizes. 2
We also analyzed data on fiscal year 2021 CDFI Program, RRP, and
ECIP awards to determine the number of institutions that received (1)
CDFI Program and RRP awards; (2) CDFI Program and ECIP awards;
and (3) CDFI Program, RRP, and ECIP awards. For each of the three
groups, we calculated the median award amount by program.
1Two types of entities that are not yet certified CDFIs are eligible to receive technical
assistance awards through the CDFI Program or Native American CDFI Assistance
Program: (1) Emerging CDFIs are noncertified entities that demonstrate in their award
applications that they have an acceptable plan to meet CDFI certification requirements by
the end of the award period of performance or another date that the CDFI Fund selects.
(2) Sponsoring entities are legal organizations that primarily serve Native communities
and plan to create a separate legal entity that will become a certified CDFI primarily
serving Native communities by a date specified in the award assistance agreement.
2Treasury had not made awards for the third emergency program—Equitable Recovery
Program—as of October 2022.
Page 50 GAO-23-105952 Community Development Financial Institutions
Appendix II: Objectives, Scope, and
Methodology
To determine how institutions have used CDFI Program and RRP funds,
we analyzed data from two types of reports that recipients submit to
Treasury via AMIS—Uses of Award Reports and Transaction Level
Reports (TLR). Uses of Award reports provide information on the amount
of award funds recipients spent on different categories of activities. We
obtained data that CDFI Program financial and technical assistance
award recipients reported to Treasury from fiscal year 2016 to June 30,
2022, and data that RRP recipients reported from the program’s inception
through June 30, 2022. The data included reports from recipients of CDFI
Program financial and technical assistance awards and recipients of RRP
awards. 3 We analyzed these data to calculate the amount and share of
award dollars that CDFI Program and RRP recipients spent on the
financial and technical assistance categories.
While the Uses of Award data specifically reflect institutions’ spending of
program funds, the TLR data include information on all of the transactions
(such as loans and investments) recipients made, regardless of funding
source. We obtained TLR data for fiscal years 2016–2020. 4 The data only
include CDFI Program recipients because RRP awards were made in
June 2021. 5 Also, because technical assistance recipients are not
required to submit these reports, the data only include financial
assistance recipients. We analyzed these data to determine the number
and dollar amount of the transactions recipients made, as well as the
transaction type (e.g., term loan, line of credit, or equity investment) and
3Institutions can receive an award from the CDFI Program for either financial assistance
or technical assistance, but not both, in the same funding round. Financial assistance
awards must be used for activities in six specific categories (such as providing financial
products or services), while technical assistance awards must be used for activities in a
different set of seven categories (such as staff compensation or travel costs). In contrast,
RRP awards can be used for activities in any of these financial or technical assistance
categories. However, there is a limitation on how much of the RRP award can be spent on
the seven categories associated with CDFI Program technical assistance awards
(recipients may spend up to $200,000 or 15 percent of their grant, whichever is greater, on
activities in the seven categories). The data we obtained did not include reports from
recipients of financial or technical assistance awards from the Native American CDFI
Assistance Program or reports on the use of supplemental financial assistance from the
CDFI Program.
4From fiscal years 2018 through 2020, Treasury required recipients to submit data only on
transactions they originated during the reporting period, while in fiscal years 2016 and
2017, it required recipients to submit data on all of the transactions that were outstanding
at any time during the reporting period. However, the data we obtained from Treasury for
fiscal years 2016 and 2017 only include transactions recipients originated during the
reporting period in order to allow for accurate comparisons across years.
5The TLR data also included recipients of financial assistance from the Native American
CDFI Assistance Program.
Page 51 GAO-23-105952 Community Development Financial Institutions
Appendix II: Objectives, Scope, and
Methodology
purpose (e.g., consumer, business, or home purchase). We also
analyzed characteristics of individuals and businesses that received
transactions, including income status, credit score, and business structure
(e.g., for-profit, nonprofit, or other structures).
To supplement this work, we also conducted a literature review to identify
and summarize research on the economic impact of CDFI Program funds.
Our review encompassed studies published from January 2010 to early
July 2022. To identify studies, we searched databases of scholarly
articles, public policy research, and Federal Reserve System publications,
among others. We identified three studies, each of which were reviewed
by a GAO economist. Two of the studies then underwent a secondary
review by a GAO methodologist. Based on these reviews, we determined
that one of the three studies provided reliable information about the
economic impact of CDFI Program funds.
Additionally, to provide insight on locations potentially affected by CDFI
Program awards, we analyzed geographic, demographic, and
socioeconomic information of the counties where recipients were located
and the census tracts where they were conducting projects. For our
county-level analysis, we used the matched CDFI Program recipient data
discussed earlier to identify the addresses of fiscal year 2019–2021
award recipients. 6 We used this information, in addition to other data sets
described later in this appendix, to determine (1) the metropolitan
classifications of recipients’ reported locations and (2) the county-level
characteristics of those locations. 7 For our census tract-level analysis, we
used census tract codes for projects reported by CDFI Program financial
assistance recipients in the TLR data for fiscal years 2016–2020. These
codes represent the census tracts associated with the individuals,
businesses, and projects that received loans and investments from the
CDFIs. Again, the TLR data included all transactions made by each CDFI,
rather than only those funded using CDFI Program awards.
To identify the metropolitan classifications of recipients’ locations, we first
used the U.S. Department of Housing and Urban Development’s ZIP
6As described earlier, we were able to match about 68 percent of fiscal year 2019–2021
CDFI Program award recipients to ACR data.
7Institutions only reported one location in their annual certification reports. For institutions
with multiple branches, the data do not include additional locations beyond the one
reported in the certification reports. CDFIs may make investments in communities outside
of their reported location.
Page 52 GAO-23-105952 Community Development Financial Institutions
Appendix II: Objectives, Scope, and
Methodology
code to county code data set to convert recipient ZIP codes to county
codes for recipients who had convertible ZIP codes. 8 We then identified
the metropolitan classifications of those recipients’ counties by matching
their county codes to the U.S. Department of Agriculture’s (USDA) 2017
metropolitan/non-metropolitan county classifications. 9
To identify the socioeconomic and demographic characteristics of
counties where recipients were located, we used the county codes we
previously matched to USDA’s county classifications and compared them
with available data from the Bureau of the Census American Community
Survey (Census ACS), 2016–2020, data, which are 5-year estimates of
county-level characteristics in the United States. We reviewed
characteristics such as mean percentage of county residents that were
non-White, mean poverty rate, and median household income, among
others. To understand differences between counties where CDFIs were
located and counties where there were no CDFIs, we compared the mean
characteristics of the two sets.
To determine the socioeconomic and demographic characteristics of the
census tracts where CDFIs conducted projects, we identified the available
census tract codes for the transactions for which we had data. We then
compared these census tract codes with available Census ACS data,
2015–2019, which are 5-year estimates of census tract-level
characteristics in the United States. We reviewed characteristics such as
mean percentage of census tract residents that were non-White, mean
percentage of census tract residents that were participating in public
assistance programs, the mean poverty rate, and median housing values,
among others. To understand differences between census tracts where
CDFIs were active and census tracts where they were not, we compared
the mean characteristics of the two sets.
We assessed the reliability of the Treasury data we used for objectives 1
and 2 by reviewing agency documentation, including Treasury data
collection instructions, and by conducting electronic tests for outliers,
missing data, and erroneous values. We also interviewed Treasury
officials to discuss the interpretation and reliability of various data fields.
8Five CDFIs were located in ZIP codes that did not have a county equivalent in the data
set we used.
9The most recent update of these classifications was in 2017. Two of the CDFIs in our
data set were located in Puerto Rico, which does not have metropolitan/non-metropolitan
county classifications, so they are not included in this analysis.
Page 53 GAO-23-105952 Community Development Financial Institutions
Appendix II: Objectives, Scope, and
Methodology
We determined the data we used were sufficiently reliable for describing
selected attributes of award recipients and transaction participants
(individual and business borrowers), award and transaction amounts,
uses of funds, and selected characteristics of areas where award
recipients were located or active.
Steps Treasury Took to To examine steps Treasury took to address risks posed by the
Address Emergency emergency programs, we examined policies and procedures for the
emergency programs on eligibility and targeting requirements, recipient
Program Risks
selection, and post-award financial terms and reporting requirements. We
also compared the policies and procedures for the emergency programs
to those for the CDFI Program in fiscal year 2021 to identify streamlined
processes that could introduce risks to program integrity. We reviewed
documentation and interviewed Treasury officials who designed the
programs to determine whether and how Treasury addressed the risks we
identified. We also reviewed the emergency programs’ policies and
procedures to determine what controls were in place to help ensure
awards were made in accordance with statutory requirements and to
institutions that had the capacity to use the funds as intended.
We also reviewed documentation on and interviewed Treasury officials
about agency mechanisms to help prevent fraud, waste, and abuse in the
emergency programs. Because the emergency programs were in the
early stages of implementation, we did not evaluate the extent to which
Treasury’s approach aligned with leading practices for fraud risk
management. Additionally, we reviewed requirements in the Payment
Integrity Information Act of 2019 regarding assessment of improper
payment risks. We reviewed the improper payment risk assessment
Treasury conducted for RRP and discussed with Treasury officials their
plans for completing assessments for the other emergency programs.
Finally, we reviewed Treasury Office of Inspector General (OIG) reports
on the emergency programs and Treasury documentation of steps taken
to address OIG recommendations.
We conducted this performance audit from March 2022 to December
2022 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.
Page 54 GAO-23-105952 Community Development Financial Institutions
Appendix III: Information on the Lines of Appendix III: Information on the Lines of
Business and Services Provided by CDFI
Business and Services Provided by CDFI
Program Recipients
Program Recipients
We analyzed available information on Community Development Financial
Institutions (CDFI) that received CDFI Program awards in fiscal years
2019–2021, including (1) their primary lines of business, (2) the
development services they reported providing, and (3) the financial
services they reported providing. 1 Over that period, about 75 percent of
CDFI recipients with available data cited consumer, residential real
estate, or business finance as their primary line of business (see table
19). About 25 percent cited commercial real estate, intermediary finance,
microfinance, or “other” as their primary business line.
Table 19: Primary Business Lines of CDFI Program Recipients with Available Data, Fiscal Years 2019–2021
Commercial Residential
Business Consumer Intermediary real estate real estate
finance finance finance Microfinance finance finance Other Total
Number 121 137 11 69 37 118 10 503
Percentage 24.1% 27.2% 2.2% 13.7% 7.4% 23.5% 2.0% 100%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: For our analysis, we matched Treasury’s fiscal year 2019–2020 certification data to
Community Development Financial Institutions (CDFI) Program awards data for fiscal years 2019–
2021. Because of the unavailability of fiscal year 2021 certification data and because not all CDFIs
who received awards were required to submit these data, this analysis includes the approximately 68
percent of recipients for which data were available.
The data in the table include institutions that received financial or technical assistance awards from
the CDFI Program or the Native American CDFI Assistance Program, as well as those that received
supplemental assistance awards for the Healthy Food Financing Initiative or Persistent Poverty
Counties.
During that same period, according to available data, about one-half of
the development services that CDFIs reported providing were technical
assistance services, such as business assistance and homeownership
counseling (see table 20). About 37 percent of services provided focused
on consumer education (i.e., credit counseling and financial education),
and the remaining 13 percent were other services.
1Fiscal year 2021 CDFI certification data were not yet available when we conducted our
analyses. For our analysis, we matched the Department of the Treasury’s fiscal year
2019–2020 certification data to CDFI Program awards data for fiscal years 2019–2021.
Our analysis does not capture all recipients because of the unavailability of fiscal year
2021 certification data and because not all CDFIs that received awards in fiscal years
2019–2021 were required to submit annual certification data (e.g., emerging CDFIs). Our
analysis includes the approximately 68 percent of recipients for which data were available.
Page 55 GAO-23-105952 Community Development Financial Institutions
Appendix III: Information on the Lines of
Business and Services Provided by CDFI
Program Recipients
Table 20: Types of Development Services CDFI Program Recipients Offered with Available Data, Fiscal Years 2019–2021
Business Homeownership Real estate
Credit Financial technical counseling and technical
counseling education assistance technical assistance assistance Other Total
Number 442 527 640 366 267 346 2,588
Percentage 17.1% 20.4% 24.7% 14.1% 10.3% 13.4% 100%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: For our analysis, we matched Treasury’s fiscal year 2019–2020 certification data to
Community Development Financial Institutions (CDFI) Program awards data for fiscal years 2019–
2021. Because of the unavailability of fiscal year 2021 certification data and because not all CDFIs
who received awards were required to submit these data, this analysis includes the approximately 68
percent of recipients for which data were available.
The data in the table include institutions that received financial or technical assistance awards from
the CDFI Program or the Native American CDFI Assistance Program, as well as those that received
supplemental assistance awards for the Healthy Food Financing Initiative or Persistent Poverty
Counties.
According to available data, CDFIs reported that loans accounted for
about 96 percent of the financial services they provided from fiscal years
2019 through 2021 (see table 21). The remainder consisted of equity
investments (about 3 percent) and loan guarantees (1 percent).
Table 21: Types of Financial Services CDFI Program Recipients Provided with Available Data, Fiscal Years 2019–2021
Equity investments Loan guarantees Loans Total
Number 76 35 2,775 2,886
Percentage 2.6% 1.2% 96.2% 100%
Source: GAO analysis of Department of the Treasury data. | GAO-23-105952
Notes: For our analysis, we matched Treasury’s fiscal year 2019–2020 certification data to
Community Development Financial Institutions (CDFI) Program awards data for fiscal years 2019–
2021. Because of the unavailability of fiscal year 2021 certification data and because not all CDFIs
who received awards were required to submit these data, this analysis includes the approximately 68
percent of recipients for which data were available.
The data in the table include institutions that received financial or technical assistance awards from
the CDFI Program or the Native American CDFI Assistance Program, as well as those that received
supplemental assistance awards for the Healthy Food Financing Initiative or Persistent Poverty
Counties.
Page 56 GAO-23-105952 Community Development Financial Institutions
Appendix IV: GAO Contact and Staff
Appendix IV: GAO Contact and Staff
Acknowledgments
Acknowledgments
William B. Shear, (202) 512-8678 or ShearW@gao.gov
GAO Contact
In addition to the contact named above, Steve Westley (Assistant
Staff Director), Melissa Kornblau (Analyst in Charge), David Ballard, Chelsea
Acknowledgments Carter, Hannah Dodd, Garrett Hillyer, Jill Lacey, John McGrail, Zamir
Ruli, and Jena Sinkfield made key contributions to this report.
(105952)
Page 57 GAO-23-105952 Community Development Financial Institutions
The Government Accountability Office, the audit, evaluation, and investigative
GAO’s Mission arm of Congress, exists to support Congress in meeting its constitutional
responsibilities and to help improve the performance and accountability of the
federal government for the American people. GAO examines the use of public
funds; evaluates federal programs and policies; and provides analyses,
recommendations, and other assistance to help Congress make informed
oversight, policy, and funding decisions. GAO’s commitment to good government
is reflected in its core values of accountability, integrity, and reliability.
The fastest and easiest way to obtain copies of GAO documents at no cost is
Obtaining Copies of through our website. Each weekday afternoon, GAO posts on its website newly
GAO Reports and released reports, testimony, and correspondence. You can also subscribe to
GAO’s email updates to receive notification of newly posted products.
Testimony
Order by Phone The price of each GAO publication reflects GAO’s actual cost of production and
distribution and depends on the number of pages in the publication and whether
the publication is printed in color or black and white. Pricing and ordering
information is posted on GAO’s website, https://www.gao.gov/ordering.htm.
Place orders by calling (202) 512-6000, toll free (866) 801-7077, or
TDD (202) 512-2537.
Orders may be paid for using American Express, Discover Card, MasterCard,
Visa, check, or money order. Call for additional information.
Connect with GAO on Facebook, Flickr, Twitter, and YouTube.
Connect with GAO Subscribe to our RSS Feeds or Email Updates. Listen to our Podcasts.
Visit GAO on the web at https://www.gao.gov.
Contact FraudNet:
To Report Fraud,
Website: https://www.gao.gov/about/what-gao-does/fraudnet
Waste, and Abuse in
Automated answering system: (800) 424-5454 or (202) 512-7700
Federal Programs
A. Nicole Clowers, Managing Director, ClowersA@gao.gov, (202) 512-4400, U.S.
Congressional Government Accountability Office, 441 G Street NW, Room 7125, Washington,
Relations DC 20548
Chuck Young, Managing Director, youngc1@gao.gov, (202) 512-4800
Public Affairs U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, DC 20548
Stephen J. Sanford, Managing Director, spel@gao.gov, (202) 512-4707
Strategic Planning and U.S. Government Accountability Office, 441 G Street NW, Room 7814,
External Liaison Washington, DC 20548
Please Print on Recycled Paper.
File and source
- File
- gao-23-105952.pdf
- Size
- 1,318,962 bytes
- SHA-256
- e4484eafc92ac5cbce7adb60ac4439d2df6bc7447dc52ae4a5be6006754e73ce
- Our copy
- gao-23-105952.pdf
- Original
- www.gao.gov