This is the accessible text file for GAO report number GAO-10-905 
entitled 'Housing and Community Grants: HUD Needs to Enhance Its 
Requirements and Oversight of Jurisdictions' Fair Housing Plans' 
which was released on October 15, 2010. 

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

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

Report to Congressional Requesters: 

United States Government Accountability Office:
GAO: 

September 2010: 

Housing and Community Grants: 

HUD Needs to Enhance Its Requirements and Oversight of Jurisdictions' 
Fair Housing Plans: 

GAO-10-905: 

GAO Highlights: 

Highlights of GAO-10-905, a report to congressional requesters. 

Why GAO Did This Study: 

Pursuant to the Fair Housing Act, Department of Housing and Urban 
Development (HUD) regulations require grantees, such as cities, that 
receive federal funds through the Community Development Block Grant 
CDBG) and HOME Investment Partnerships Program (HOME) to further fair 
housing opportunities. In particular, grantees are required to prepare 
planning documents known as Analyses of Impediments (AI), which are to 
identify impediments to fair housing (such as restrictive zoning or 
segregated housing) and actions to overcome them. HUD has oversight 
responsibility for AIs. This report (1) assesses both the conformance 
of CDBG and HOME grantees AIs’ with HUD guidance pertaining to their 
timeliness and content and their potential usefulness as planning 
tools and (2) identifies factors in HUD’s requirements and oversight 
that may help explain any AI weaknesses. 

GAO requested AIs from a representative sample of the nearly 1,200 
grantees, compared the 441 AIs received (95 percent response based on 
final sample of 466) with HUD guidance and conducted work at HUD 
headquarters and 10 offices nationwide. 

What GAO Found: 

On the basis of the 441 AIs reviewed, GAO estimates that 29 percent of 
all CDBG and HOME grantees’ AIs were prepared in 2004 or earlier, 
including 11 percent from the 1990s, and thus may be outdated. HUD 
guidance recommends that grantees update their AIs at least every 5 
years. GAO also did not receive AIs from 25 grantees, suggesting that, 
in some cases, the required documents may not be maintained, and 
several grantees provided documents that did not appear to be AIs 
because of their brevity and lack of content. GAO reviewed 60 of the 
current AIs (those dating from 2005 through 2010) and found that most 
of these documents included several key elements in the format 
suggested in HUD’s guidance, such as the identification of impediments 
to fair housing and recommendations to overcome them. (See table below 
for common impediments identified in 30 of these 60 current AIs.) 
However, the vast majority of these 60 AIs did not include time frames 
for implementing their recommendations or the signatures of top 
elected officials, as HUD guidance recommends, raising questions about 
the AI’s usefulness as a planning document. As a result, it is unclear 
whether the AI is an effective tool for grantees that receive federal 
CDBG and HOME funds to identify and address impediments to fair 
housing. 

HUD’s limited regulatory requirements and oversight may help explain 
why many AIs are outdated or have other weaknesses. Specifically, HUD 
regulations do not establish requirements for updating AIs or their 
format, and grantees are not required to submit AIs to the department 
for review. A 2009 HUD internal study on AIs, department officials, 
and GAO’s work at 10 offices identified critical deficiencies in these 
requirements. For example, HUD officials rarely request grantees’ AIs 
during on-site reviews to assess their compliance with overall CDBG 
and HOME program requirements, limiting the department’s capacity to 
assess AIs’ timeliness and content. While HUD initiated a process to 
revise its AI regulatory requirements in 2009, what the rule will 
entail or when it will be completed is not clear. In the absence of a 
department-wide initiative to enhance AI requirements and oversight, 
many grantees may place a low priority on ensuring that their AIs 
serve as effective fair housing planning tools. 

Table: Commonly Cited Impediments to Fair Housing in Selected AIs: 

Impediment: Zoning and site selection
Description of impediment: Building and zoning codes, which may 
contain lot requirements such as minimum street frontage and front 
yard setbacks, and amenities (e.g., landscaping), that can affect the 
feasibility of developing low- and moderate-income housing. 

Impediment: Public services
Description of impediment: Inadequate public services, such as 
schools, in areas where minorities or people with disabilities may 
live. 

Impediment: Lending policies and practices
Description of impediment: Less favorable mortgage lending terms from 
private lenders, such as higher interest rates, for minority borrowers 
than are generally available for nonminority borrowers with similar 
risk characteristics. 

Source: GAO analysis of 30 current AIs. 

[End of table] 

What GAO Recommends: 

GAO recommends that, through regulation, HUD require grantees to 
update their AIs periodically, follow a specific format, and submit 
them for review. HUD neither agreed nor disagreed with the 
recommendations but noted recent efforts to improve compliance and 
oversight. 

View [hyperlink, http://www.gao.gov/products/GAO-10-905] or key 
components. For more information, contact Orice Williams Brown at 
(202) 512-8678 or williamso@gao.gov. 

[End of section] 

Contents: 

Letter: 

Background: 

Many Grantees' AIs Are Outdated or Otherwise Out of Conformance with 
HUD Guidance and Thus There Is Limited Assurance They Serve as 
Effective Fair Housing Planning Tools: 

HUD's Limited Regulatory Requirements and Oversight and Enforcement 
Approaches May Help Explain Why Many AIs Are Outdated or Contain Other 
Weaknesses: 

Conclusions: 

Recommendations for Executive Action: 

Agency Comments and Our Evaluation: 

Appendix I: Objectives, Scope, and Methodology: 

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

Appendix III: GAO Contact and Staff Acknowledgments: 

Tables: 

Table 1: HUD's Suggested Format for AI: 

Table 2: Authorship of 60 Current AIs: 

Table 3: Most Commonly Cited Impediments to Fair Housing in Selected 
AIs: 

Table 4: HUD's Key Elements Represented in 60 Current AIs: 

Table 5: CPD On-site Reviews in 2009 for Selected HUD Field Offices 
and Number of Times Grantees AIs Were Reviewed or Obtained: 

Table 6: Reasons Grantees Offered for Not Sending AIs to GAO: 

Figures: 

Figure 1: Estimated Completion Dates of CDBG and HOME Program 
Grantees' AIs: 

Figure 2: Estimated Percentages of Outdated AIs, by HUD Region: 

Abbreviations: 

AFFH: affirmatively further fair housing: 

AI: Analysis of Impediments to Fair Housing Choice: 

CAPER: Consolidated Annual Performance and Evaluation Report: 

CDBG: Community Development Block Grant: 

ConPlan: Consolidated Plan: 

CPD: Office of Community Planning and Development: 

DCI: data collection instrument: 

ESG: Emergency Shelter Grants: 

FHAct: Fair Housing Act of 1968: 

FHEO: Office of Fair Housing and Equal Opportunity: 

HMDA: Home Mortgage Disclosure Act: 

HOME: HOME Investment Partnerships Program: 

HOPWA: Housing Opportunities for Persons with AIDS: 

HUD: Department of Housing and Urban Development: 

Recovery Act: American Recovery and Reinvestment Act of 2009: 

[End of section] 

United States Government Accountability Office:
Washington, DC 20548: 

September 14, 2010: 

Congressional Requesters: 

The Fair Housing Act (FHAct) of 1968, as amended, prohibits 
discrimination in the sale, rental, and financing of housing on the 
basis of race or color, religion, sex, national origin, familial 
status, or disability.[Footnote 1] Section 808(e) (5) of the FHAct 
also requires the Department of Housing and Urban Development (HUD) to 
administer its housing and urban development programs, including 
formula grant programs such as the Community Development Block Grant 
program (CDBG) and HOME Investment Partnerships Program (HOME), in a 
manner that affirmatively furthers fair housing (AFFH)--that is, that 
eliminates housing discrimination, promotes fair housing choice, and 
fosters compliance with the nondiscrimination provisions of the FHAct. 
[Footnote 2] In fiscal year 2009, these programs provided over $5 
billion in funding to nearly 1,200 eligible communities and 
jurisdictions and the median grant size was about $1.4 million. To 
help fulfill the AFFH requirement, HUD regulations require grant 
recipients, such as municipalities and counties, to: 

* prepare a planning document known as an Analysis of Impediments to 
Fair Housing Choice (AI), which is to identify any potential 
impediments to fair housing choice within the jurisdiction (such as 
exclusive zoning laws or segregated housing patterns); 

* take appropriate actions to overcome the effects of any impediments 
identified in the AI, such as revising zoning ordinances; and: 

* maintain records reflecting the analysis and actions taken. 

HUD views the AI as a tool that serves as the foundation for fair 
housing planning by municipalities, counties, and other grantees that 
receive federal funds through their participation in the CDBG, HOME, 
and other grant programs. The department is also responsible for 
overseeing and enforcing grantees' compliance with AFFH requirements 
and guidance, including those pertaining to their AIs. 

Grantees' compliance with AFFH requirements and the effectiveness of 
HUD's oversight and enforcement have been called into question through 
litigation and reports and testimonies for some years. In August 2009, 
the Department of Justice and HUD announced a fair housing settlement 
with Westchester County in New York, a CDBG grantee that was required 
to conduct an AI. A federal district court had concluded that the 
county was aware that racial and ethnic segregation and discrimination 
persisted in its municipalities, but its AI made no mention of these 
practices or any plan to address them. The litigation ended in an 
agreement that required the county to invest $51.6 million in 
affordable housing over the next 7 years and to undertake and fund 
marketing, public education, and other outreach efforts to promote 
fair and affordable housing.[Footnote 3] Further, fair housing groups 
have conducted studies and stated in testimony that many grantees fail 
to prepare substantive AIs and that HUD's oversight and enforcement of 
the AI requirements have been minimal.[Footnote 4] In 2009, HUD 
completed an internal study, which found that many AIs were outdated 
or appeared to have been prepared in a cursory fashion and found that 
the department's oversight was limited.[Footnote 5] Associations that 
represent grantees, such as the U.S. Conference of Mayors, have stated 
that their members are committed to furthering fair housing but have 
noted a variety of challenges involved in doing so, including limited 
funding for conducting the AIs and a lack of clear HUD guidance on 
preparing an AI. HUD officials have stated that the department has not 
always fulfilled its obligation to assist grantees in meeting AFFH 
requirements but is in the process of developing a rule that will 
provide the guidance grantees need.[Footnote 6] 

Given your questions about grantees' compliance with AFFH 
requirements, particularly the quality of their AIs, and HUD's 
oversight activities, you asked that we conduct a review of relevant 
issues. Specifically, our report (1) assesses both the conformance of 
CDBG and HOME grantees' AIs with HUD guidance pertaining to their 
timeliness and content, as well as the AIs' potential usefulness as 
fair housing planning tools, and (2) identifies factors that may help 
explain any potential weaknesses in grantees AIs, particularly factors 
related to HUD's regulatory requirements and oversight and enforcement 
approaches. 

To address the first objective, we selected a representative sample of 
473 CDBG and HOME grantees from the total population of 1,209 fiscal 
year 2009 program participants and requested their most recent AIs. We 
received AIs from 441 of the 466 grantees in our final sample, for a 
response rate of 95 percent.[Footnote 7] We reviewed these 441 AIs to, 
among other things, estimate the extent to which all CDBG and HOME 
program grantees had updated their AIs in accordance with HUD guidance 
and an internal department study on AI compliance.[Footnote 8] 
Further, we analyzed a nonrepresentative subset of 60 of the 281 
current AIs we received (those prepared from 2005 through 2010 that 
were among the total of 441 AIs provided) to determine whether 
grantees or an external party had prepared the documents, and compared 
them with HUD guidance to determine whether they followed the 
department's recommended format. Finally, we reviewed 30 AIs from this 
nonrepresentative subset to categorize commonly identified types of 
potential impediments to fair housing choice.[Footnote 9] 

For the second objective, we reviewed HUD's policies, procedures, and 
guidance for overseeing and enforcing the AFFH requirement, 
particularly pertaining to AIs, as well as information on resource 
levels for these activities. We also reviewed the 2009 internal HUD 
study on AI compliance and oversight; annual reports that grantees are 
required to submit to HUD that, among other items, may discuss AFFH 
activities; public testimony by senior HUD officials; and 
documentation and data on enforcement activity. In addition, we 
interviewed HUD officials at headquarters and in three regional and 
seven field offices that we selected on the basis of varying criteria, 
including the size of the jurisdiction, as measured by the number of 
grantees within each jurisdiction; geographic location; and potential 
risk of nonconformance, as measured by the estimated incidence of 
outdated or missing AIs. Appendix I provides more details about our 
objectives, scope, and methodology. 

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

Background: 

A number of HUD's large grant programs are subject to the AFFH 
requirement. These programs include the following: 

* CDBG program: The CDBG program is authorized by Title I of the 
Housing and Community Development Act of 1974, as amended. The program 
provides annual grants to states, metropolitan areas, and urban 
counties to fund an extensive array of community development 
activities, such as providing decent housing and a suitable living 
environment and expanding economic opportunities that primarily 
benefit Americans of modest financial means. The CDBG program was 
funded at about $3.6 billion in fiscal year 2009, making CDBG the 
largest grant program, with the largest number of grantees. 

* HOME program: The HOME program is the largest government-sponsored 
affordable housing production program. HOME was authorized under Title 
II of the Cranston-Gonzalez National Affordable Housing Act of 1990, 
as amended, and provides grants to states and localities, often in 
partnership with local nonprofit groups. These grants are used to fund 
a wide range of activities that build, buy, and rehabilitate 
affordable housing for rent or sale and provide direct rental 
assistance to low-income people. The program was funded at about $1.8 
billion in fiscal year 2009. 

* HOPWA program: HOPWA is intended to address the urgent housing needs 
of low-income Americans living with HIV/AIDs, who are 
disproportionately represented in low-income minority communities. 
HOPWA funds may be used for a wide range of housing, social services, 
program planning, and development costs. HOPWA funds also may be used 
for health care and mental health services, chemical dependency 
treatment, nutritional services, case management, assistance with 
daily living, and other supportive services. HOPWA was funded at about 
$310 million in fiscal year 2009. 

* ESG program: The ESG provides homeless persons with basic shelter 
and essential supportive services. It also provides short-term 
homeless prevention assistance to persons at imminent risk of losing 
their own housing due to eviction, foreclosure, or utility shutoffs. 
The ESG was funded at about $1.7 billion, in fiscal year 2009. 

AI Requirement Established in HUD Regulations: 

To help ensure that grantees receiving funds through the CDBG and 
other formula grant programs are meeting the AFFH requirements, HUD 
regulations require them to prepare and maintain AIs. HUD defines the 
AI as a comprehensive review of potential impediments and barriers to 
the right to be treated fairly when seeking housing. The AI is 
expected to cover public and private policies, practices, and 
procedures affecting housing choice and assess how they all affect the 
location, availability, and accessibility of housing. Grantees are 
also to develop strategies and actions to overcome these barriers 
based on history, circumstances, and experiences. In effect, the AI is 
a tool that is intended to serve as the basis for fair housing 
planning; provide essential information to policymakers, 
administrative staff, housing providers, lenders, and fair housing 
advocate; and assist in building public support for fair housing 
efforts. Grantees may use a portion of their CDBG and other grant 
funds to prepare their AIs, and AIs may be prepared by the grantees 
themselves or under contract with external parties, such as fair 
housing groups, consultants, universities, or others. 

While HUD regulations require grantees to prepare AIs, other 
requirements pertaining to these local planning documents are limited. 
For example, HUD has not issued regulations specifying how often 
grantees should update their AIs or the specific elements that should 
be included in them. HUD regulations also do not require grantees to 
submit their AIs to the department for review and approval. Instead, 
CDBG and HOME grantees are to annually certify to HUD that they are 
meeting AFFH requirements, which include having prepared an AI, taking 
steps to address identified impediments, and maintaining records of 
their actions. HUD generally accepts grantees annual AFFH 
certifications, including that they have prepared AIs, and will not 
initiate further reviews unless evidence to the contrary emerges from 
complaints or through the department's routine monitoring activities. 

While HUD has not issued regulations that specifically define when the 
grantees' must update their AIs, or what elements they must include, 
it has issued recommended guidance on these subjects. As discussed in 
this report, for example, HUD has issued guidance recommending that 
grantees update their AIs every 3 to 5 years. In 1996, HUD also issued 
a fair housing guide, which included a suggested format for AIs and 
other important fair housing planning elements (see table 1). The 
format and elements include an introduction and executive summary; 
jurisdictional and background data, such as demographic data and 
analysis; and an evaluation of the jurisdiction's current fair housing 
legal status, such as a listing and description of fair housing 
related complaints that have been filed and their status or 
resolution. Further, the suggested format and other elements include a 
listing of the impediments identified, proposed actions and time 
frames to overcome them, and the signatures of top elected officials. 

Table 1: HUD's Suggested Format for AI: 

Suggested element: Introduction and executive summary of the analysis; 
Description: Explains who conducted the AI and identifies the 
participants and methodology used, funding source, and summaries of 
impediments found and actions to address them. 

Suggested element: Jurisdictional background data; 
Description: Includes demographic, income, employment, housing 
profile, maps, and other relevant data. 

Suggested element: Evaluation of jurisdiction's current fair housing 
legal status; 
Description: Discusses fair housing complaints and compliance reviews 
that have resulted in a charge or finding of discrimination, fair 
housing discrimination suits filed by the Department of Justice or 
private plaintiffs, the reasons for any trends or patterns in 
complaints and enforcement, and other fair housing concerns. 

Suggested element: Identification of impediments to fair housing 
choice; 
Description: Identifies impediments to fair housing. 

Suggested element: Conclusions and recommendations for overcoming 
impediments; 
Description: Summarizes any impediments identified in the analysis and 
presents recommendations to overcome identified impediments. 

Suggested element: Time frames for implementing actions to overcome 
impediments; 
Description: Sets out the time frame for completing each action or set 
of actions to serve as milestones toward achieving the actions. 

Suggested element: Signature page; 
Description: Includes the signature of a chief elected official, such 
as a mayor. 

Source: HUD 1996 Fair Housing Planning Guide. 

Note: HUD's 1996 Fair Housing Planning Guide suggested format included 
one other element, description of current public and private fair 
housing programs and activities in the jurisdiction, which was not 
included in our review. While the fair housing guidance's suggested AI 
format does not include time frames for implementing recommendations 
to address identified impediments, time frames are discussed elsewhere 
in the guide as a component of fair housing planning. 

[End of table] 

Grantees' AFFH Activities May Also Be Addressed in Other Required 
Reports to HUD: 

Although HUD does not require grantees to submit their AIs to the 
department for review and approval, it does require them to 
periodically submit other reports on their overall use of CDBG and 
other grant program funds, such as HOME, HOPWA, and ESG. In some 
cases, HUD specifically requires that grantees include in these 
reports information about their AFFH activities. These AFFH activity 
reports include the following: 

* Consolidated Plan (ConPlan): The ConPlan, which grantees must file 
with HUD for review and approval every 5 years, is a planning document 
that identifies low-and moderate-income housing needs within a 
community and specifies how grantees intend to use federal funds to 
address those needs. According to HUD, the purpose of the ConPlan is 
to enable grantees to shape the various housing and community 
development programs into effective, coordinated neighborhood and 
community development strategies. Within their ConPlans, grantees are 
to provide their AFFH certifications annually, as described here. 

* Annual Action Plan: Grantees are required to submit an Annual Action 
Plan. These annual plans lay out how the grantees plan to achieve the 
overall objectives in their consolidated plans in the coming fiscal 
year. 

* Consolidated Annual Performance and Evaluation Report (CAPER): 
Within 90 days at the end of the program year, grantees that have 
approved ConPlans must file a CAPER with HUD, which reports on the 
progress they have made in carrying out the activities described in 
their Annual Action Plans. The CAPER must include, among other things, 
actions taken to affirmatively further fair housing. HUD is 
responsible for reviewing the accuracy of the CAPER. Because HUD does 
not require grantees to submit AIs, the CAPER serves as the main 
document that department officials use to learn about grantees' fair 
housing activities and accomplishments. 

HUD's AFFH Oversight Structure Involves Two Offices and Numerous 
Regional and Field Offices: 

Two HUD offices share responsibility for overseeing CDBG and HOME 
grantees' compliance with AFFH requirements, including those 
pertaining to their AIs: the Office of Community Planning and 
Development (CPD) and the Office of Fair Housing and Equal Opportunity 
(FHEO). CPD is responsible for helping ensure that grantees are in 
overall compliance with CDBG and other grant program requirements. For 
example, CPD is responsible for ensuring that grantees spend federal 
funds on approved activities, such as affordable housing creation and 
community development. To carry out their oversight activities, CPD 
staff are to review and approve grantees' ConPlans, Annual Action 
Plans, and CAPERs and conduct on-site monitoring reviews of a limited 
sample of high-risk grantees each year to assess their compliance with 
various program requirements, including those pertaining to AFFH. HUD 
has the authority to disapprove a ConPlan if the grantee's AFFH 
certification is inaccurate or missing. Disapproval of a ConPlan may 
result in withholding CDBG and other formula grant funds until the 
grantee submits an adequate AFFH certification within an established 
time frame. While CPD serves as HUD's main liaison with grantees, FHEO 
maintains final authority to determine and resolve matters involving 
fair housing compliance, including the AFFH requirement. In carrying 
out their responsibilities, FHEO staff may use the results of CPD's 
oversight activities, such as its reviews of grantees' reports and on- 
site monitoring reviews. FHEO staff also are to independently review 
grantee reports, such as their CAPERs, and may conduct on-site 
monitoring reviews of grantees on a limited basis. 

HUD maintains 10 regional offices and 81 field offices. CPD and FHEO 
staff are located in approximately 44 of the field offices, which have 
primary responsibility for monitoring and enforcing AFFH requirements, 
including those pertaining to AIs. Field and regional office staff 
generally consult with CPD and FHEO officials in HUD's headquarters 
offices on key decisions and activities, including whether to 
disapprove a grantee's ConPlan or AI. 

Many Grantees' AIs Are Outdated or Otherwise Out of Conformance with 
HUD Guidance and Thus There Is Limited Assurance They Serve as 
Effective Fair Housing Planning Tools: 

While we estimate that the majority of grantees have current AIs in 
accordance with HUD guidance, many others may be outdated per the 
guidance. Specifically, we estimate that 29 percent of all AIs were 
prepared in 2004 or earlier, including 11 percent that date from the 
1990s. Because many grantees' AIs are outdated, they may not provide a 
reliable basis to identify and mitigate current impediments to fair 
housing that may exist within their communities. We also (1) did not 
receive AIs from 25 grantees despite repeated requests that they 
provide them, which suggests that, in some cases, grantees may not 
maintain the documents as is required; and (2) several grantees 
provided documents with their status as AIs not clear due to their 
brevity and lack of content. While the majority of grantees may have 
current AIs, we question the usefulness of many such AIs as fair 
housing planning documents. We reviewed a subset of current AIs we 
received (those dating from 2005 through 2010) for a variety of 
reasons, including to gain insights into the types of impediments they 
identified and to determine whether they included the key elements 
identified by HUD in its 1996 fair housing guidance. [Footnote 10] The 
most commonly cited impediments to fair housing choice were zoning 
restrictions, inadequate public services in low-and moderate-income 
areas, lending discrimination, and a lack of public awareness about 
fair housing rights. Further, we found that current AIs generally 
contained several basic elements suggested in HUD's guidance, such as 
demographic data and analysis, and recommendations to overcome 
identified impediments. However, a significant majority of the current 
AIs did not identify time frames for implementing the recommendations 
or contain the signatures of top elected officials as is also 
suggested in HUD's guidance. As a result, these AIs may not provide a 
reliable basis for measuring the grantees' progress in overcoming 
impediments or reasonable assurance that top elected officials endorse 
the recommendations in the AI and are accountable for implementing 
them. In sum, our review found limited assurances that grantees are 
placing needed emphasis on preparing AIs as effective planning tools 
to identify and address potential impediments to fair housing as 
required by statutes governing the CDBG and HOME programs and HUD 
regulations and guidance.[Footnote 11] 

An Estimated 29 Percent of AIs Are Outdated and Thus May Not Provide a 
Reliable Basis for Identifying and Mitigating Current Impediments to 
Fair Housing: 

We estimate that while 64 percent of all grantees have current AIs, 29 
percent may be outdated having been prepared in 2004 or earlier 
(including 11 percent from the 1990s), and the date for 6 percent 
could not be determined (fig. 1). While HUD has not officially defined 
what constitutes an outdated AI through regulation, it has issued 
guidance that addresses how often an AI should be updated. Using HUD's 
guidance and interviews with department officials as criteria, we 
define an AI as outdated if it was completed in 2004 or earlier. 
Specifically, HUD's 1996 Fair Housing Planning Guide--the main 
reference document for grantees in developing AIs--suggests that 
grantees conduct or update them at least every 3 to 5 years, in part 
to be consistent with the consolidated planning cycle.[Footnote 12] On 
February 14, 2000 and again on September 2, 2004, HUD issued 
memorandums to all CPD and FHEO officials to remind grantees to update 
their AIs annually when necessary, but especially at the beginning of 
a new consolidated 5-year planning cycle. In addition, HUD's 2009 
study on grantees' AI conformance concluded that the grantees with AIs 
dating from the 1990s may place a low priority to them and that such 
AIs should be updated.[Footnote 13] 

Figure 1: Estimated Completion Dates of CDBG and HOME Program 
Grantees' AIs: 

[Refer to PDF for image: horizontal bar graph] 

AIs from: Date unknown: 6%. 

AIs from: 1990s thru 2004: 29%; 
1990s: 11%; 
2000-2004: 18%. 

AIs from: 2005 to 2010: 64%. 

Source: GAO analysis of GAO data. 

Note: Percentages do not add to 100 due to rounding. 

[End of figure] 

The incidence of outdated AIs was generally consistent across the 
country and among both large and small grantees.[Footnote 14] Figure 2 
shows the percentage of grantees with outdated AIs in each of the 
geographic areas covered by HUD's 10 regional offices, which ranged 
from a low of 14 percent in Region IV to a high of 45 percent in 
Region VII. Despite the variation, only one region had a statistically 
significant difference between its percentage of outdated AIs and the 
percentage at the national level.[Footnote 15] 

Figure 2: Estimated Percentages of Outdated AIs, by HUD Region: 

[Refer to PDF for image: U.S. map, depicting HUD Regions] 

Region 1: 25%; 
Region 2: 44%; 
Region 3: 29%; 
Region 4: 14%; 
Region 5: 36%; 
Region 6: 33%; 
Region 7: 45%; 
Region 8: 30%; 
Region 9: 28%; 
Region 10: 23%. 

Sources: GAO analysis of GAO data; Art Explosion (map). 

[End of figure] 

HUD's 2009 internal study on AI conformance also concluded that many 
grantees' AIs were outdated.[Footnote 16] Specifically, in examining 
the timeliness of 45 AIs in its sample, HUD found that about 18 
percent (8) were produced before 2000 and had not yet been updated. 
While the HUD study provides some insights into the AIs, its findings 
cannot be generalized to the entire population because of limitations 
in its sampling methodology.[Footnote 17] 

Because many grantees' AIs are outdated, they may not provide a 
reliable basis for identifying and mitigating impediments to fair 
housing. For example, HUD's 1996 fair housing guidance suggests that 
grantees use demographic data from the U.S. Census Bureau when 
preparing their AIs and to update these planning documents as new 
census data becomes available. Updated census data could indicate 
demographic trends within a jurisdiction that might be useful in 
preparing an AI, such as whether particular areas of a jurisdiction 
are becoming progressively more or less segregated over time and the 
potential reasons thereof. Moreover, according to one FHEO field 
office official, grantees should update their AIs every 5 years per 
the guidance, because the impediments to fair housing in a particular 
community evolve and change, and new issues can occur on a continuing 
basis. Another FHEO field office official said that an AI dated from 
the 1990s would not be considered current under any circumstances. The 
official said that grantees should update their AIs periodically to 
adjust to the development of potential impediments to fair housing 
choice within their communities. For example, the official noted that, 
subprime mortgage lending grew substantially during the 2000s, and 
subprime mortgage lenders potentially disproportionately targeted 
minority borrowers, which resulted in many foreclosures among such 
groups.[Footnote 18] Without taking steps to update their AIs, whether 
grantees that receive federal funds through the CDBG, HOME, and other 
grant programs are sufficiently focused on overcoming current 
impediments to fair housing that may exist within their communities is 
unclear. 

We Did Not Receive AIs from 25 Grantees and the Brevity and Lack of 
Content of Others Raised Questions about Whether They Were AIs: 

We did not receive AIs from 25 grantees, despite intensive follow up 
efforts that included multiple e-mails and phone calls to appropriate 
officials. Representatives from some of these grantees offered several 
reasons for not providing the requested AIs. For example, 
representatives from two grantees said that they could not find their 
AIs, and a representative from another said an AI had not been 
prepared. Further, representatives from 8 grantees stated that they 
had already sent us their AIs, although we have no record of receiving 
them. We cannot definitively determine that all these grantees are out 
of compliance with statutes and HUD regulations requiring grantees to 
maintain AIs. However, the failure of these grantees to provide AIs, 
together with the results of HUD's 2009 study that also found that 
some grantees did not provide AIs as requested, raises questions about 
whether some jurisdictions may be receiving federal funds without 
preparing the documents required to demonstrate that they have taken 
steps to affirmatively further fair housing.[Footnote 19] 

Our analysis of the 441 AIs we received from grantees also indicates 
that the documents ranged in length from several hundred pages with 
supporting graphs and other materials to a few pages of content. 
[Footnote 20] For example, one grantee's AI contained 64 tables 
illustrating a wide variety of information, ranging from a breakdown 
of the grantee's population by race, ethnicity, and poverty status to 
the rates at which low-income applicants in the grantee were denied 
conventional loans. While AIs may consist of many pages, length does 
not necessarily indicate the quality of these documents. For example, 
some lengthy AIs we reviewed had reports attached, including their 
CAPERS, that grantees were required to submit to HUD separately. On 
the other end of the spectrum, we identified five documents whose 
status as AIs was unclear based on their brevity and limited content. 
Specific examples are as follows: 

* One grantee provided a four-page survey of residents within the 
community on fair housing issues. 

* One grantee provided a two-page document that largely discussed its 
progress in implementing a local statute pertaining to community 
preservation and that contained two sentences describing a fair 
housing impediment. 

* One grantee provided a three-page document that contained 
descriptions of activities designed to help the homeless and other 
special needs groups and described the actions that the grantee took 
to address barriers to affordable housing. 

* One grantee provided a four-page description of the community 
itself, and it did not identify impediments to fair housing. 

* One grantee provided a two-page e-mail that identified one 
impediment to fair housing choice, and in follow up conversations an 
official from this grantee, confirmed that the document constituted 
its AI. 

Given the brevity and lack of content in these documents, they may not 
constitute AIs as required by the CDBG and HOME statutes and HUD 
regulations. 

Analysis of Current AIs Indicates That Authorship and Types of 
Identified Impediments to Fair Housing Vary Widely: 

While many AIs are outdated or, in some cases, grantees may not 
maintain the documents as required by HUD, we did estimate that 64 
percent of grantees have prepared current AIs. To gain insights into 
current AIs, we reviewed a subset of 60 of the 281 such documents that 
we received for a variety of reasons, including to determine their 
authorship.[Footnote 21] HUD's 1996 guidance suggests that AIs may be 
authored by grantees, fair housing or industry groups, universities or 
colleges, or any combination thereof. Our analysis indicates that 
grantees, through their community development and planning offices, 
for example, had prepared about half of the 60 current AIs we reviewed 
(table 2). In 9 of the 60 cases we reviewed, the grantee had 
contracted with a fair housing organization; in 9 of the 60 cases with 
a private consulting firm; and in 4 cases with a university or 
college. In about 6 of the cases, the AI did not identify the author, 
or the author's identity was not clear. 

Table 2: Authorship of 60 Current AIs: 

Author type: Grantee; 
Number: 29. 

Author type: Consulting firm; 
Number: 9. 

Author type: Fair housing organization; 
Number: 9. 

Author type: University or college; 
Number: 4. 

Author type: Mixed (more than one author type); 
Number: 2. 

Author type: Other (e.g., housing task force); 
Number: 1. 

Author type: Cannot determine author; 
Number: 6. 

Author type: Total; 
Number: 60. 

Source: GAO analysis of 60 AIs that were prepared from 2005 through 
2010. 

[End of table] 

We also reviewed 30 of the 60 current AIs to identify the types of 
impediments to fair housing choice that had been identified by the 
grantees.[Footnote 22] While these grantees cited a variety of 
potential impediments in their AIs, at least half identified four 
types of impediments (table 3): (1) zoning and site selection, (2) 
neighborhood revitalization, (3) lending policies and practices, and 
(4) fair housing informational programs.[Footnote 23] 

Table 3: Most Commonly Cited Impediments to Fair Housing in Selected 
AIs: 

Impediments: Zoning and site selection; 
Description of impediments: 
* Building and zoning codes, which may contain lot requirements such 
as minimum street frontage and front yard setbacks, and amenities 
(e.g., landscaping), that can affect the feasibility of developing low-
and moderate-income housing; 
* Placement of new or rehabilitated housing for low-and moderate-
income groups in areas that already have high concentrations of this 
type of housing or have zoning requirements that encourage such 
concentrations. 

Impediments: Neighborhood revitalization, municipal and other 
services, employment-housing transportation linkage; 
Description of impediments: 
* Inadequate public services in low-and moderate-income areas, where 
many African-American, Hispanics, and people with disabilities may 
live, including schools, recreational facilities, social service 
programs, parks, roads, transportation, street lighting, trash 
collection, and police protection. 

Impediments: Lending policies and practices; 
Description of impediments: 
* Less favorable mortgage lending terms from private lenders, such as 
higher interest rates, for African-Americans or other minority 
borrowers than are generally available for nonminority borrowers with 
similar risk characteristics. 

Impediments: Informational programs; 
Description of impediments: 
* Lack of access to information about the rights and responsibilities 
associated with fair housing, potentially creating an environment 
favorable to discriminatory practices. 

Source: GAO analysis of 30 current AIs. 

[End of table] 

We also identified specific examples of each of these four impediments 
and the grantees' planned actions to address them as described here. 

* Zoning and site selection. One AI we reviewed, which was prepared on 
behalf of several grantees by a regional planning unit within a local 
university, identified some of their established or planned land use 
policies as potential impediments to fair housing. For example, the AI 
found that some of the grantees had minimum lot-size requirements for 
building single-family residences that could limit housing 
affordability. The AI noted that a 1-acre minimum lot size, for 
example, would create land costs that would make owning or renting 
homes on such lots unaffordable to low-income families. Further, the 
AI noted that some of the grantees were considering requiring that all 
new homes be constructed of brick, a requirement that could 
substantially increase construction costs compared with siding and 
make the homes unaffordable to low-income families. To address these 
potential impediments, the AI recommended that the grantees (1) ensure 
that a sufficient portion of their communities were zoned for 
multifamily construction and that lot sizes for single-family housing 
were small enough to keep single-family housing affordable and (2) 
consider the potential impact on housing affordability, including on 
minority families, before adopting building codes that require all-
brick construction. 

* Neighborhood revitalization, municipal and other services, 
employment-housing transportation linkage. One AI prepared by a 
private nonprofit fair housing organization on behalf of several 
grantees noted that the area's transportation system was inadequate to 
service the needs of all residents. The AI concluded that residents 
who wanted or needed to use public transportation were obliged to 
limit their residences to the jurisdictions in which their jobs were 
located even if they wanted to live elsewhere. To address this 
impediment, the AI recommended that the grantees support a regional 
transportation system that not only provided services to low-and 
moderate-income households throughout the area but also met the needs 
of employers in geographic areas that were not currently served. 

* Lending policies and practices. An AI for a large county comprising 
several grantees included a review of data required under the Home 
Mortgage Disclosure Act (HMDA).[Footnote 24] The analysis found that 
mortgage lenders in the jurisdiction denied the applications of upper- 
income black applicants at a rate that was three times higher than the 
rate for equally situated white applicants. Further, the AI found that 
the loan denial rate for Hispanic mortgage applicants was twice as 
high as that of equally situated white applicants. The AI recommended 
that the grantees contract with a consultant to prepare and conduct 
training for mortgage lenders to encourage their voluntary compliance 
with fair housing laws. Finally, the AI recommended that the grantees 
continue to monitor HMDA data to determine if the educational programs 
had a positive effect on loan denial rates for minorities. 

* Fair housing informational programs. An AI for a county concluded 
that, because the grantee received few complaints from residents about 
fair housing, there might be a lack of public knowledge about fair 
housing rights and responsibilities. The AI suggested that because 
residents might not be aware of such issues, landlords and others 
involved in the real estate business could feel that they had more 
leeway in dealing with potential home buyers and renters. As a result, 
the AI recommended that the grantee should promote fair housing 
education through public workshops, presentations at schools and 
libraries, public service announcements in English and Spanish, and 
the distribution of fair housing literature at all county facilities 
and events. 

Lack of Time Frames for Implementing Recommendations and the 
Signatures of Top Elected Officials Raises Questions about the 
Usefulness of Many Current AIs as Planning Documents: 

While current AIs may identify a variety of potential impediments to 
fair housing and strategies to overcome them, questions exist about 
the status of many such AIs as local planning documents. As part of 
our review, we found that the 60 current AIs generally included five 
of the seven key elements in the suggested format for AIs contained in 
HUD's 1996 fair housing guidance (table 4). Specifically, four of 
these elements were present in over 55 of the 60 grantees' AIs: 
jurisdictional background data, evaluation of fair housing legal 
status, identifications of impediments to fair housing choice, and 
conclusions and recommendations. The introduction and executive 
summary were present in 52 of the grantees' AIs. However, we found 
that only 12 of the AIs included time frames for implementing 
recommendations for overcoming impediments and that only 8 AIs 
included the signatures of top elected officials. 

Table 4: HUD's Key Elements Represented in 60 Current AIs: 

Suggested elements to be included in the AI: Introduction and 
executive summary of the analysis; 
Number: 52. 

Suggested elements to be included in the AI: Jurisdictional background 
data; 
Number: 57. 

Suggested elements to be included in the AI: Evaluation of 
jurisdiction's current fair housing legal status; 
Number: 56. 

Suggested elements to be included in the AI: Identification of 
impediments to fair housing choice; 
Number: 58. 

Suggested elements to be included in the AI: Conclusions and 
recommendations for overcoming impediments; 
Number: 56. 

Suggested elements to be included in the AI: Time frames for 
implementing actions for overcoming impediments; 
Number: 12. 

Suggested elements to be included in the AI: Signature page; 
Number: 8. 

Source: GAO analysis of a subset of 60 randomly selected current AIs. 

Note: HUD's 1996 Fair Housing Planning Guide suggested format included 
one other element, description of current public and private fair 
housing programs and activities in the jurisdiction, which was not 
included in our review. While the fair housing guidance's suggested AI 
format does not include time frames for implementing recommendations 
to address identified impediments, time frames are discussed elsewhere 
in the guide. 

[End of table] 

The lack of time frames for implementing proposed actions among a 
substantial majority--48 of the subset of 60 current AIs we reviewed-- 
is potentially significant. HUD's guidance on including estimated time 
frames for implementing recommendations is generally consistent with 
our view that time frames are an important component of effective 
strategic and other planning process. Recognizing the importance of 
specific time frames, officials from one HUD field office we contacted 
routinely provided technical advisory notices to aid grantees that 
were preparing or updating their AIs. These notices recommend, among 
other things, that grantees include benchmarks and timetables for 
implementing actions in their AIs. In the absence of established time 
frames in AIs, determining whether grantees are achieving progress in 
implementing recommendations to overcome identified impediments is 
difficult. 

Further, our finding that 52 of the 60 current AIs did not include the 
signatures of top elected officials raises questions as to whether the 
officials endorse the analyses and support suggested actions in the 
AIs and are accountable for implementing them. These questions may be 
particularly significant with respect to the 25 AIs identified in 
table 2 that were prepared by an external party under contract, such 
as a fair housing group or consultant, rather than by the grantee 
through one of its agencies. Our review indicates that none of these 
25 AIs had been signed by the grantees' top elected officials. While 
HUD field office officials we contacted said the lack of these 
signatures did not necessarily mean that the grantees did not plan to 
implement the actions described in the documents, other HUD officials 
disagreed. For example, officials from two field offices said that the 
lack of such signatures suggested that the grantees may not endorse 
the analysis and recommendations in the AIs. Officials from one of 
these field offices said that, in the absence of the signature of a 
top elected official, an AI had little value as a planning document. 
We note that HUD requires authorizing grantee officials to sign 
documents that, among other things, certify that their ConPlans 
identify community development and housing needs, contain specific 
short-and long-term objectives to address such needs, and certify that 
the grantee is following its department-approved plan. This is not an 
uncommon accountability model for compliance-based regulatory 
structures.[Footnote 25] Without the signatures of top elected 
officials, it is not clear that grantees have established plans to 
identify and address impediments to fair housing within their 
jurisdictions. 

We identified an example of an AI that lacked time frames for 
addressing impediments and had not been signed by a top elected 
official and that did not appear to be functioning as an effective 
planning document. This AI, which was prepared under contract by a 
fair housing group, had 15 specific recommendations, including 
conducting fair lending "testing" and establishing an effective code 
enforcement program.[Footnote 26] We contacted the CDBG representative 
for this grantee, who said that he believed that the AI contained just 
two recommendations. The official also stated that, due to other 
priorities, the grantee had not yet had time to implement either of 
these two recommendations and did not have immediate plans to do so. 

HUD's 2009 study also raises questions about the usefulness of many 
AIs as planning documents to identify and address potential 
impediments to fair housing. As discussed previously, this study 
concluded that many of the AIs in its sample dated from the 1990s, 
which HUD said indicated that these grantees place a low priority on 
the documents. According to the HUD study, moreover, many of the AIs 
reviewed did not conform to the department's guidance and appeared to 
have been prepared in a cursory fashion. In sum, our findings that 
many AIs are outdated, may not be prepared as required, or lack time 
frames and signatures, together with the findings of HUD's study, 
raise significant questions as to whether the AI is effectively 
serving as a tool to help ensure that all grantees are committed to 
identifying and overcoming potential impediments to fair housing 
choice as required by statutes governing the CDBG and HOME programs 
and HUD regulations. 

HUD's Limited Regulatory Requirements and Oversight and Enforcement 
Approaches May Help Explain Why Many AIs Are Outdated or Contain Other 
Weaknesses: 

HUD's AI requirements and oversight and enforcement approaches have 
significant limitations that likely contribute to our findings that 
many such documents are outdated or contain other weaknesses. In 
particular, HUD's regulations have not established standards for 
updating AIs or the format that they must follow, and grantees are not 
required to submit their AIs to the department for review. According 
to HUD's 2009 internal report on AI compliance, and CPD and FHEO 
officials, the limited regulatory requirements pertaining to AIs and 
limited resources and competing priorities adversely affect the 
department's capacity to help ensure the effectiveness of AIs as fair 
housing planning documents. Moreover, our work involving 10 HUD field 
offices identified specific instances that illustrate the limitations 
in the department's AI oversight and enforcement approaches and the 
need for corrective actions.[Footnote 27] For example, we found that 
HUD officials rarely request grantees' AIs during on-site monitoring 
reviews or receive complaints from the public about such documents, 
which means that the department often has minimal information about 
the status of grantees' AIs in terms of their timeliness and content. 
Conversely, while we identified instances where certain field offices 
took proactive steps to help ensure the integrity of the AI process, 
such as one office's efforts to better ensure that grantees update 
their AIs periodically, these initiatives were not common. Recognizing 
the limitations in its AI requirements and oversight and enforcement 
approaches, in 2009, HUD initiated a process to update relevant 
regulations, but it is not clear what issues any revised regulatory 
requirements will address or when they will be completed. HUD has also 
developed plans to address limited staffing resources that may have 
undermined its capacity to oversee grantees' AIs or implement any new 
regulatory initiatives, but it is unclear how effective the 
initiatives will be. We note that some proposals that have been made, 
such as a requirement that grantees submit their AIs to HUD for 
review, would not necessarily involve a significant commitment of 
staff resources and could have important benefits. In the absence of a 
department-wide initiative to strengthen AI requirements and oversight 
and enforcement, many grantees may place a low priority on ensuring 
that their AIs serve as effective planning tools. 

HUD Regulations Do Not Establish Standards for the Timeliness of AIs 
or their Format, and Grantees Are Not Required to Submit Them for the 
Department's Review: 

As discussed previously, HUD's regulatory requirements pertaining to 
AIs are limited. While HUD regulations require grantees to prepare 
AIs, they do not specify when grantees must update them or the 
specific format they must follow. Moreover, HUD's regulations do not 
require grantees to submit their AIs to the department on a routine 
basis for review to help ensure their effectiveness as a tool to 
identify and address impediments to fair housing. Instead, pursuant to 
statutes governing the CDBG and HOME program, grantees are required to 
annually self-certify by attesting to HUD that they are in compliance 
with the department's AFFH requirements, including those pertaining to 
the AI. Specifically, the self-certification, which is generally a one-
page document, attests that the grantee has completed an AI, has taken 
steps to overcome the impediments identified in the AI, and maintains 
records of its efforts. In general, HUD officials, pursuant to 
department regulations, are to accept these self-certifications as 
sufficient evidence that the grantee has an AI and is acting to 
implement its recommendations. 

While HUD does not require grantees to submit their AIs on a routine 
basis, CPD and FHEO officials, who share AFFH oversight and 
enforcement responsibilities, use several approaches to monitor their 
overall compliance with AFFH requirements, including those 
requirements that pertain to AIs. Specifically, these approaches, 
while limited, can involve CPD or FHEO officials obtaining grantees' 
AIs and following up as may be deemed necessary. The following efforts 
describe how HUD generally carries out these responsibilities: 

* Reviews of grantee reports and plans that, among other things, are 
to address AFFH compliance. HUD CPD and FHEO officials are to 
regularly review documents that grantees annually submit on their 
overall plans and performance in complying with CDBG and other grant 
program requirements. For example, at the end of the fiscal year, 
grantees are required to submit their CAPER to HUD, which discusses 
their progress in meeting their objectives for the use of CDBG and 
other grant funds. As part of the CAPER, HUD requires grantees to 
include a description of actions taken to AFFH. If determined 
necessary by a HUD reviewer of either the Annual Action Plan or the 
CAPER, the department could request that a grantee provide its AI for 
review and analysis. 

* On-site monitoring reviews to assess grantee compliance with HUD 
requirements, which can include reviews of AFFH documentation, such as 
AIs. Under HUD policy, CPD field officials are to conduct a limited 
number of risk-based, on-site monitoring reviews each year to assess 
grantees compliance with a variety of CDBG and other grant 
requirements. In some cases, FHEO officials may join CPD officials on 
these monitoring reviews or conduct independent monitoring reviews. 
HUD headquarters establishes annual criteria for assessing risk each 
year and the percentage of on-site monitoring reviews to be conducted. 
The criteria can include the amount of the CDBG or HOME grant, the 
amount of time that has passed since the last on-site review, and 
employee turnover in grantee offices responsible for implementation of 
CDBG and other grant programs. CPD officials generally were directed 
to visit at least 10 to 15 percent of the grantees under their 
jurisdiction annually. As part of these reviews, CPD may request that 
grantees provide copies of their most recent AIs. CPD staff or FHEO 
staff may review these AIs and follow up with the grantees where 
deemed warranted. 

* Reviews as part of a complaint. HUD may also receive complaints 
about grantees' AFFH compliance from a range of sources, including 
individuals, fair housing groups, or federal, state, or local 
agencies. In conducting investigations in response to such complaints, 
CPD or FHEO staff may request that grantees provide their AIs as 
deemed appropriate. 

If HUD officials identify concerns with grantees AIs through these 
processes, they can take several different actions. These actions 
include: 

* technical assistance, such as training workshops, to complete an AI; 

* a "Special Assurance" document, which HUD may draft in order to 
outline a number of tasks that a grantee must do to fulfill 
requirements, including describing actions to overcome the effects of 
identified impediments, and creating a timetable for accomplishing 
these actions; these assurances usually are signed by the grantee's 
chief elected official to signify cooperation; and: 

* withholding CDBG and HOME funding by disapproving a grantee's 
ConPlan for failure to comply with requirements, including completion 
of an AI. According to HUD officials, this action is a last resort and 
rarely used. 

HUD's 2009 Study, CPD and FHEO Officials and Our Work at 10 Field 
Offices Have Identified Limitations in the Department's Capacity to 
Help Ensure That AIs Serve as Effective Planning Tools: 

HUD's 2009 internal study, CPD and FHEO officials in headquarters and 
field offices, as well as our analysis involving 10 field offices have 
identified significant limitations in the department's long-standing 
AI requirements and oversight and enforcement approaches. HUD 
officials also cited staffing resource constraints as undermining 
their oversight capacity and ability to implement corrective measures. 
Regarding HUD's 2009 study, it concluded that the department's limited 
AI regulatory requirements and oversight processes contributed to the 
study's findings that many AIs were outdated or otherwise did not 
conform with the department' 1996 fair housing guidance. To better 
ensure that grantees conform with HUD guidance, the report suggested 
requiring grantees to submit their AIs for review and approval. The 
report noted that, because grantees are not currently required to 
submit AIs to HUD, a possible first step could simply be to implement 
a submission requirement. However, the report also noted that HUD 
would have to dedicate sufficient resources to conduct reviews of AIs 
and develop appropriate criteria for assessing them. The study 
suggested that HUD consider both (1) requiring grantees to post their 
AIs on the Internet and (2) compiling all submitted AIs to be posted 
online at a single clearing house Web site to enhance transparency and 
increase public awareness of the documents. Further, the study 
suggested that HUD update its fair housing guidance and provide 
additional technical assistance to grantees to help ensure they 
prepare more effective AIs. For example, the study suggested that HUD 
assist grantees in obtaining the data necessary to prepare AIs and 
provide relevant training. However, HUD officials said the department 
has not yet acted to implement the recommendations in the study. 

While HUD has not yet acted on the recommendations in the 2009 
internal study, senior headquarters officials cited limited regulatory 
requirements as adversely affecting oversight efforts. In the absence 
of specific regulatory requirements, CPD officials said it is 
difficult for field offices to ensure that grantees update their AIs 
within specified time frames or conform to a specific format in 
preparing the documents, including the signatures of top elected 
officials. In contrast, the CPD staff noted that there are specific 
regulatory requirements pertaining to grantees ConPlans, Annual Action 
Plans, and CAPERs, including when these documents must be prepared and 
what must be included, which facilitates their oversight efforts. 
Additionally, CPD and FHEO officials in HUD's headquarters cited 
limited resources and competing regulatory priorities as limiting 
oversight of grantees' AIs and AFFH compliance generally and 
potentially posing challenges to any new regulatory initiatives. CPD 
officials said that obtaining and reviewing grantees' AIs is a low 
priority for field office staff due to competing demands and limited 
resources, and additional resources and technical expertise would be 
required for staff to review and approve AIs as suggested in the 
department's internal study. 

CPD and FHEO officials from the 10 HUD field offices we contacted also 
commented about limited AFFH regulatory requirements and oversight 
approaches. For example, officials from 7 of the 10 field offices told 
us that because grantees are not required to submit their AIs, 
verifying whether grantees had AIs or had updated them was difficult. 
Some field office staff officials also said that, because grantees are 
not required to submit their AIs, their capacity to assess grantees' 
overall compliance with AFFH requirements is limited. For example, 
without requiring grantees to submit their AIs, the officials said 
that they could not verify whether the potential impediments to fair 
housing choice that may be cited in other documents, such as the 
grantees' 5-year ConPlan and CAPERs, were the same impediments listed 
in their AIs. One field office officials suggested that HUD require 
grantees to submit their AIs as part of their 5-year consolidated 
plans to enable them to verify that the two documents were consistent. 
Officials from several field offices also recommended that HUD revise 
its regulations to require that AIs meet certain standards for 
timeliness and completeness, which they said would enhance their 
abilities to oversee and enforce the program. 

Further, CPD and FHEO field office officials agreed with HUD 
headquarters officials that declining resources and competing 
priorities had limited their ability to assess grantees' AIs or AFFH 
compliance generally. Representatives from all of the 10 HUD offices 
we contacted said that their staff levels had decreased recently while 
their workload had increased, especially with the implementation of 
the American Recovery and Reinvestment Act of 2009 (Recovery Act). 
[Footnote 28] For example, FHEO and CPD officials in several field 
offices told us that they were losing staff due to retirements and 
promotions. One FHEO Field Office Director said that it had one 
official currently available to monitor 54 entitlement grantees' 
compliance with all relevant statutes and regulations, including those 
pertaining to AFFH, within the office's jurisdiction. Additionally, 
CPD and FHEO officials in one office commented that, at one time they 
had enough staff to regularly send both a CPD and FHEO representative 
on on-site monitoring reviews, but with staff reductions over the 
years, they are no longer able to continue this practice. Field office 
officials also stated that work priorities were often shifting, making 
it difficult for them to consistently focus on one aspect of CDBG and 
other grant program compliance. 

We identified the following specific examples that illustrate how 
HUD's limited oversight and enforcement program on a nationwide basis 
may have contributed to many AIs being outdated or not otherwise in 
full conformance with department guidance, and further support the 
need for corrective actions to address these limitations: 

* Our review of CAPERs for a group of grantees with outdated AIs 
raises questions about the value of such reports as a means of 
assessing AFFH compliance. As discussed previously, HUD's annual 
reviews of CAPERs and other required grantee reports are a key means 
by which department officials assess AFFH compliance in the absence of 
a requirement that grantees routinely submit their AIs for review. We 
selected a nongeneralizable sample of 30 grantees with outdated AIs 
from the 441 grantees that sent us AIs.[Footnote 29] We requested that 
HUD provide the most recently available CAPER report for each of these 
30 grantees to help us determine what information these reports 
contain about the grantees' AIs, and the department provided 27 
CAPERs. In 17 of the 27 cases, the grantees mentioned that they had an 
AI but did not specify the AI's date. In such cases, HUD field offices 
that rely on CAPER reviews to help assess AFFH compliance may not be 
aware AIs are outdated unless they specifically follow up with the 
grantee to find out the date of its AI. In 10 cases, the grantees' 
CAPERs disclosed the date of their AI which, in some cases, was from 
the 1990s. The extent to which HUD officials identify such disclosures 
in CAPERs or follow up on them was not clear. 

* Field offices' CPD on-site monitoring programs and complaint review 
processes provide a limited basis for assessing grantees' AIs and 
taking follow up actions as may be required. We obtained data from 7 
of the 10 field offices we contacted regarding the number of CPD on-
site grantee reviews they conducted in 2009 and the number of times 
they obtained or reviewed AIs during such monitoring. As table 5 
indicates, CPD field office officials reported collecting or reviewing 
AIs in 17 of the 88 reviews. Moreover, officials from these 7 field 
offices said they rarely if ever receive public complaints about a 
grantee's AI. Given the absence of public complaints, which could be 
due to the fact that there may be a general public unawareness that 
grantees are required to prepare AIs, the complaint process does not 
appear to provide a systematic basis for HUD to identify potential 
limitations in grantees' AIs and follow up with them as necessary. 

Table 5: CPD On-site Reviews in 2009 for Selected HUD Field Offices 
and Number of Times Grantees AIs Were Reviewed or Obtained: 

HUD office: Office 1; 
Number of grantees: 23; 
Number of CPD on-site grantee monitoring reviews: 2; 
Number of AI's that are reviewed and/or obtained: 0. 

HUD office: Office 2; 
Number of grantees: 54; 
Number of CPD on-site grantee monitoring reviews: 32; 
Number of AI's that are reviewed and/or obtained: 0. 

HUD office: Office 3; 
Number of grantees: 21; 
Number of CPD on-site grantee monitoring reviews: 19; 
Number of AI's that are reviewed and/or obtained: 2. 

HUD office: Office 4; 
Number of grantees: 18; 
Number of CPD on-site grantee monitoring reviews: 6; 
Number of AI's that are reviewed and/or obtained: 2. 

HUD office: Office 5; 
Number of grantees: 25; 
Number of CPD on-site grantee monitoring reviews: 3; 
Number of AI's that are reviewed and/or obtained: 3. 

HUD office: Office 6; 
Number of grantees: 29; 
Number of CPD on-site grantee monitoring reviews: 4; 
Number of AI's that are reviewed and/or obtained: 4. 

HUD office: Office 7; 
Number of grantees: 105; 
Number of CPD on-site grantee monitoring reviews: 22; 
Number of AI's that are reviewed and/or obtained: 6. 

HUD office: Total; 
Number of grantees: 275; 
Number of CPD on-site grantee monitoring reviews: 88; 
Number of AI's that are reviewed and/or obtained: 17. 

Source: GAO analysis of CPD monitoring data from seven field offices. 

[End of table] 

* Our visits to field offices located in the two HUD regions with the 
highest incidence of outdated AIs illustrate some of the inherent 
limitations in the department's oversight and enforcement processes. 
According to an FHEO official in one of these offices, historically, 
it has generally been unaware of the current state of a grantee's AIs, 
because the grantees did not routinely submit them. In 2010, this 
official said the field office requested that all of the grantees 
under its jurisdiction submit their AIs, so that the office could gain 
a better perspective on the timeliness of the AIs. In one case, the 
official said that the office learned that a grantee that had been 
certifying its AFFH compliance for several years did not have an AI. 
At the other field office, we identified instances where it accepted 
AFFH certifications under questionable circumstances. For example, in 
one case, representatives from a grantee told us that they could not 
find its AI. When we asked field office officials about this 
circumstance, they said that the grantee had provided a two-page 
summary of an AI that it completed in 1996. Field office officials 
said that they viewed the summary as sufficient evidence that the 
grantee had completed an AI. Moreover, in distributing CDBG and other 
grant funds each year, this field office sends routine communications 
which, among other things, remind grantees to update their AIs 
periodically. While this is a potentially positive step, the field 
office does not appear to take any additional steps to ensure that AIs 
are not outdated. For example, we identified at least one grantee 
subject to the field office's jurisdiction that has received such 
communications for four consecutive years, but its AI was prepared in 
the 1990s. 

While our analysis generally verified that there are limitations in 
HUD's overall AI oversight and enforcement approaches that require 
corrective action, we identified practices in certain field offices 
that appear designed to better ensure that AIs are effective planning 
documents and that grantees fulfill overall AFFH requirements. As 
discussed previously, in 2010, one HUD field office we contacted 
independently requested that all grantees within its jurisdiction 
provide their AIs for review, which allowed FHEO officials to 
determine that many such AIs were outdated, and one grantee did not 
have an AI. In another example, officials from one field office said 
that they maintain ongoing communications with grantees to, among 
other things, determine the date they completed their AIs. In cases 
where an AI is determined to be outdated, officials told us that they 
work closely with the grantee through technical assistance to bring 
the AI up to date. We corroborated the field office's assertion by 
reviewing the AIs we received from grantees under its jurisdiction and 
found that all of the grantees had sent us updated AIs. Another field 
office has established procedures to use special assurance agreements, 
which were discussed previously, to help ensure that grantees revise 
AIs that may have identified deficiencies. However, these initiatives 
by individual field offices appear to be isolated examples within 
HUD's general approach to AI and AFFH oversight, which provides 
limited assurances that AIs serve as effective planning tools to 
identify and address impediments to fair housing. 

HUD Has Taken Steps to Begin to Address Regulatory, Oversight, and 
Resource Limitations, but These Efforts Are Ongoing: 

Recognizing the limitations in AI and AFFH requirements and its 
oversight processes, in 2009, HUD initiated a process to review and 
revise its existing regulations. In 2009, HUD officials held several 
"listening sessions" with key stakeholders, such as grantees and fair 
housing groups, to help identify approaches to enhance the AI and AFFH 
processes. In January 2010, HUD's Assistant Secretary for Fair Housing 
and Equal Opportunity testified that the department was working on a 
proposed regulation to enhance AFFH compliance.[Footnote 30] According 
to a senior HUD attorney, revising the AFFH regulation is a priority 
for HUD, and the proposed rule may cover a variety of topics, 
including enhancements to the guidance provided to grantees on 
preparing AIs and improvements in the department's oversight and 
enforcement approaches. However, until the rule is proposed, it is not 
clear what topics it will address. The HUD attorney also said that the 
department's tentative time frame for publishing a proposed rule to 
revise AFFH requirements is in December of 2010. However, the attorney 
also said that this proposed time frame had not been finalized and was 
subject to change. 

HUD has also established initiatives to help address staffing 
limitations that, as discussed previously, may have affected its 
overall CDBG and other grant program oversight and enforcement 
approaches, including those pertaining to AIs and AFFH requirements, 
as well as its capacity to implement any new regulatory initiatives. 
For example, on March 26, 2010, the HUD Secretary sent out a 
memorandum on the agency's Targeted Recruitment Strategy for fiscal 
years 2010-2012. In this document, the Secretary described a strategy 
for addressing HUD's need to identify qualified individuals for its 
talent pipeline over the next three fiscal years. The Secretary stated 
that this strategy would incorporate the utilization of various 
federal programs that are designed to recruit and retain students to 
positions in the federal government, such as the Presidential 
Management Fellows Program, the Student Career Experience Program; and 
the Student Temporary Employment Program. During our review, a CPD 
official said that the office recently announced a buyout for certain 
officials and that CPD was "moving more aggressively to recruit and 
hire approximately 50 new employees in the next 3 months with the 
skills to provide grant oversight, assess grantee and community 
needs," among other activities. 

While the effects of these plans and initiatives remain to be seen, we 
note that some of the proposals to enhance HUD's AI and AFFH oversight 
and enforcement approaches would not necessarily involve a significant 
commitment of additional staff resources. In particular, requiring 
grantees to submit their AIs for review, without necessarily approving 
them, would allow CPD and FHEO officials to perform a variety of basic 
tasks to better ensure their quality. These tasks could include 
verifying whether AIs (1) have been prepared as required, (2) updated 
in accordance with HUD guidance, (3) include all elements suggested in 
the 1996 fair housing guidance, and (4) are consistent with AFFH 
discussions in other key documents, particularly CAPERs. If HUD 
officials identified any areas of concern with grantees' AIs through 
such analysis, they could follow up as necessary, through technical 
assistance, enforcement actions, or other activities as may be 
necessary, to better ensure that AI serve as effective tools to 
identify and overcome impediments to fair housing. Moreover, the 
resource demands could also be mitigated if grantees submitted their 
AIs on a periodic basis over a period of time rather than all at once 
within a specified period. 

Conclusions: 

While HUD regulations have required the preparation of AIs for many 
years, whether they serve as an effective tool for grantees that 
receive federal funds through the CDBG and other programs to identify 
and address impediments to fair housing within their jurisdictions is 
unclear. We estimate that 29 percent of all AIs are outdated, 
including 11 percent that were prepared in the 1990s. Given that many 
AIs are outdated, they do not likely serve as effective planning 
documents to identify and address current potential impediments to 
fair housing choice. Moreover, some grantees may not prepare AIs, and 
others sent us cursory documentation as their AIs which, on the basis 
of their content, do not appear to be AIs. While we estimate that 64 
percent of grantees have prepared current AIs, the usefulness of many 
such AIs as planning documents is uncertain. Our review of a subset of 
60 current AIs indicates that, while many of them identify potential 
impediments to fair housing choice and contain recommendations to 
overcome them, the vast majority also lack time frames for 
implementing identified recommendations or the signatures of top 
elected officials, both of which are necessary to establish clear 
accountability to carrying out the AFFH intent. Without time frames, 
judging a grantees' progress in overcoming identified impediments is 
difficult and, without the signatures of top elected officials, 
determining whether responsible officials endorse the recommendations 
in the AIs and are accountable for ensuring their implementation is 
unclear. Absent any changes in the AI process, they will likely 
continue to add limited value going forward in terms of eliminating 
potential impediments to fair housing that may exist across the 
country. 

HUD's limited approach to establishing AI regulatory requirements, and 
its limited oversight and enforcement approaches, may help explain the 
various weaknesses in the documents that we have identified. Beyond 
requiring grantees to prepare AIs, and certify annually that they have 
done so and are addressing identified impediments, HUD requirements 
with respect to AIs are minimal. Specifically, grantees are not 
required through regulation to update their AIs periodically, include 
certain information, follow a specific format in preparing AIs, or 
submit them to HUD for review. These limitations are not new or 
unknown to HUD officials, yet little progress has been made to address 
them. While HUD officials said that the department is working on a 
regulation to enhance grantees' compliance with AFFH requirements 
since 2009, what the regulation will ultimately entail, or when it 
will be completed is unclear. In the meantime, grantees will continue 
to have considerable flexibility in determining when to update their 
AIs and what information to include in them, which could lead to 
continued weaknesses in these fair housing planning documents. 

We recognize that HUD faces resource challenges and competing 
priorities in carrying out its overall CDBG and other grant program 
responsibilities, including those pertaining to AFFH and AIs. However, 
depending on how any changes are structured, resources could be better 
leveraged to provide more coverage for overseeing grantees. For 
example, while HUD officials expressed concerns about the resources 
and technical expertise necessary to approve AIs, a grantee submission 
requirement itself could have several significant benefits without 
necessarily involving a significant commitment of staff resources. 
Specifically, a submission requirement could allow HUD staff to verify 
basic items, such as whether grantees have prepared AIs as required, 
and whether such AIs have been updated and conform to an established 
format, and are consistent with other critical reports, such as 
CAPERs. Moreover, a submission requirement would provide enhanced 
incentives for grantees to better ensure that their AIs serve as 
effective planning tools to identify potential impediments to fair 
housing and to overcome them. Failure to require that grantees submit 
their AIs on a regular basis will likely continue to result in many 
grantees not updating the documents in a timely manner or adhering to 
any guidance or requirements. 

Recommendations for Executive Action: 

To better ensure that grantees' AIs serve as an effective tool for 
grantees to identify and address impediments to fair housing, we 
recommend that HUD expeditiously complete its new regulation 
pertaining to the AFFH requirements. In so doing, we also recommend 
that HUD address three existing limitations. First, we recommend that 
HUD establish standards for grantees to follow in updating their AIs 
and the format that they should follow in preparing the documents. 
Second, to facilitate efforts to measure grantees' progress in 
addressing identified impediments to fair housing and to help ensure 
transparency and accountability, we recommend, as part of the AI 
format, HUD require grantees to include time frames for implementing 
recommendations and the signatures of responsible officials. And 
finally, we recommend HUD require, at a minimum, that grantees submit 
their AIs to the department on a routine basis and that HUD staff 
verify the timeliness of the documents, determine whether they adhere 
to established format requirements, assess the progress that grantees 
are achieving in addressing identified impediments, and help ensure 
the consistency between the AIs and other required grantee reports, 
such as the CAPERs. 

Agency Comments and Our Evaluation: 

We provided a draft of this report to HUD for its review and comment. 
We received written comments from HUD's Assistant Secretary for Fair 
Housing and Equal Opportunity, which are reprinted in appendix II. HUD 
also provided technical comments, which we have incorporated as 
appropriate. 

In its written comments, HUD highlighted its recent actions to 
affirmatively further fair housing. Among the recent initiatives to 
enhance its AFFH compliance and oversight cited in HUD's written 
comments were the following: 

* HUD's renewed commitment to AFFH. In fiscal year 2010, HUD said it 
had strengthened and clarified the AFFH requirements for grantees that 
are not specifically exempt in the FY 2010 Notice of Funding 
Availability and General Section. According to HUD, to meet this 
requirement, applicants for funding through these programs must now 
address how their proposed activities will help overcome impediments 
to fair housing choice as outlined in relevant AIs. HUD stated that by 
establishing AFFH as a policy priority within HUD's discretionary 
funding programs, will allow the department to encourage grantees to 
undertake comprehensive and innovative strategies to affirmatively 
further fair housing. For example, HUD may take steps to reward 
grantees for participating in regional efforts to promote integration 
or decreasing the concentration of poverty. 

* Increasing the level of AFFH review and technical assistance. 
According to HUD, since the beginning of fiscal year 2010, FHEO 
regional and field offices have increased the level review of 
grantee's AIs within their jurisdictions. The letter also stated that 
some offices have requested grantees to submit their AIs whereas other 
offices are doing so on a risk basis. Moreover, HUD stated that, since 
January 2010, it has increased its training of grantees regarding 
their AFFH compliance requirement. This training covers the AI and the 
importance of its completion, what information should be included in 
the AI including race and ethnicity data, who should complete the AI, 
the consequences of not completing the AI, and how to report fair 
housing activities. 

* Improving HUD's capacity for monitoring and enforcing AFFH 
compliance. HUD stated that it is exploring ways to find greater 
efficiencies and reduce staff time spent on routine administrative 
matters and to dedicate more time to AFFH oversight. This is to 
include additional training for HUD on AFFH oversight. Specifically, 
HUD said it is designing training at its National Fair Housing Academy 
on reviewing submissions to better ensure consistent and valid review 
criteria. HUD said it is also developing uniform standards for its 
staff of review of grantees' Annual Action Plans and Consolidated 
Plans for compliance with AFFH certifications. 

While we commend HUD for recognizing the need to take steps to improve 
its oversight of AFFH compliance, many of the key challenges we found 
in our report do not appear to be addressed by its current plans. 
Specifically, we note that HUD did not address the status of its 
planned AFFH rulemaking efforts, including standards for grantees to 
follow in updating their AIs and the format that they should follow in 
preparing the documents, such as including the time frames for 
implementing recommendations and the signature of responsible 
officials. Further, HUD did not discuss any plans to require, at a 
minimum, that grantees submit their AIs to the department on a routine 
basis to help ensure grantees compliance with requirements and 
guidance pertaining to these documents. In the absence of such 
regulatory requirements, the usefulness of requiring AIs as a tool to 
affirmatively further fair housing is diminished. 

As agreed with your offices, unless you publicly announce the contents 
of this report earlier, we plan no further distribution until 30 days 
from the report date. At that time, we will send copies of this report 
to the appropriate congressional committees, to the Secretary of 
Housing and Urban Development, and other interested parties. The 
report also will be available at no charge on the GAO Web site at 
[hyperlink, http://www.gao.gov]. 

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

Signed by: 

Orice Williams Brown: 
Director, Financial Markets and Community Investment: 

List of Congressional Requesters: 

The Honorable Barney Frank:
Chairman:
Committee on Financial Services:
House of Representatives: 

The Honorable Melvin L. Watt:
Chairman:
Subcommittee on Domestic Monetary Policy and Technology:
Committee on Financial Services:
House of Representatives: 

The Honorable Luis V. Gutierrez:
Chairman:
Subcommittee on Financial Institutions and Consumer Credit:
Committee on Financial Services:
House of Representatives: 

The Honorable Maxine Waters:
Chairwoman:
Subcommittee on Housing and Community Opportunity:
Committee on Financial Services:
House of Representatives: 

The Honorable Gregory W. Meeks:
Chairman:
Subcommittee on International Monetary Policy and Trade:
Committee on Financial Services:
House of Representatives: 

The Honorable Dennis Moore:
Chairman:
Subcommittee on Oversight and Investigations:
Committee on Financial Services:
House of Representatives: 

The Honorable Joe Baca:
The Honorable Michael E. Capuano:
The Honorable André Carson:
The Honorable Emanuel Cleaver:
The Honorable Keith Ellison:
The Honorable Al Green:
The Honorable Rubén Hinojosa:
The Honorable Paul W. Hodes:
The Honorable Carolyn B. Maloney:
The Honorable Carolyn McCarthy:
The Honorable Gwen Moore:
House of Representatives: 

[End of section] 

Appendix I: Objectives, Scope, and Methodology: 

The objectives of our report are to (1) assesses both the conformance 
of Community Development Block Grant (CDBG) and HOME Investment 
Partnerships Program (HOME) grantees' Analysis of Impediments (AI) 
with Department of Housing and Urban Development (HUD) guidance 
pertaining to their timeliness and content as well as the AIs' 
potential usefulness as fair housing planning tools and (2) identify 
factors that may help explain any potential weaknesses in grantees 
AIs, particularly factors related to HUD's regulatory requirements and 
oversight and enforcement approaches. 

To address the first objective, we made a document request of a 
representative sample of CDBG and HOME grantees asking that they 
submit their most recent AI to us. Although there are four HUD formula 
grant programs to which the AFFH documentation requirements apply, our 
work focused on CDBG and HOME, the two largest of such programs as 
measured by grant amount. 

Prior to launching the AI document request, we obtained contact 
information on CDBG and HOME grantees from HUD's fiscal year 2009 CDBG 
program contacts Web site.[Footnote 31] We verified that the most up- 
to-date contact information was the HUD's Office of Community Planning 
and Development (CPD) Community Connections, the clearinghouse of 
information for CPD. From the total population of 1,209 fiscal year 
2009 program participants on HUD's Web site, which includes all 1,209 
CDBG grantees and 97 percent (634 of 650) of all HOME grantees, we 
selected a random sample of 473 CDBG and HOME grantees.[Footnote 32] 
Using a two-way stratification, we stratified the population by HUD's 
10 regions and grantee's grant size (less than $500,000 and $500,000 
or more). We independently selected a random sample of 48 grantees 
from each of 10 HUD regions (with the exception of Region VII where 
there were only 41 grantees in the population). 

In January 2010, we sent out an initial e-mail to all 473 identified 
officials and requested that grantees provide their most recent AIs to 
us. To ensure a high response rate, we e-mailed follow-up requests to 
nonrespondents approximately 2 and 4 weeks after the initial e-mail. 
As a result of this follow-up, we learned that 7 grantees in our 
initial sample were out of the scope of our study and subsequently 
excluded them, thereby reducing our sample to 466 grantees.[Footnote 
33] We then conducted intensive follow up with the remaining 
nonrespondents, making repeated attempts to acquire the requested AIs 
through multiple phone calls and e-mails conducted by contractors 
hired specifically for this phase of the document request effort. 
Despite repeated attempts to follow up with nonrespondents, 25 
grantees did not submit an AI (see table 6 for rationales provided by 
officials from these 25 grantees). 

Table 6: Reasons Grantees Offered for Not Sending AIs to GAO: 

Reason cited for not sending GAO an AI: Could not find the AI; 
Number of grantees: 2. 

Reason cited for not sending GAO an AI: Never prepared an AI; 
Number of grantees: 1. 

Reason cited for not sending GAO an AI: In the process of developing 
an AI; 
Number of grantees: 2. 

Reason cited for not sending GAO an AI: Regional HUD Office has the 
AI, and the grantee's office did not; 
Number of grantees: 1. 

Reason cited for not sending GAO an AI: Claimed to have sent an AI, 
but we have not yet received it to date; 
Number of grantees: 8. 

Reason cited for not sending GAO an AI: Asked us to call back but did 
not send us an AI; 
Number of grantees: 2. 

Reason cited for not sending GAO an AI: Wrong grantee contact 
information; 
Number of grantees: 4. 

Reason cited for not sending GAO an AI: Grantees did not respond to 
phone calls and e-mails over a 2-month period; 
Number of grantees: 5. 

Reason cited for not sending GAO an AI: Total; 
Number of grantees: 25. 

Source: GAO analysis of grantees' response to our request for their 
AIs. 

[End of table] 

Upon conclusion of the document request effort, we received AIs from 
441 grantees for a response rate of 95 percent.[Footnote 34] Following 
is a summary of information from which we obtained the response rate: 

* Total population of AI: 1,209: 

* Sample size: 473: 

* Number of nonrespondents: 25: 

* Out-of-scope respondents: 7: 

* Estimated in-scope total population: 1,190: 

* In-scope respondents: 441: 

* Response rate: 94.6 percent. 

We conducted this AI request from January to March 2010. 

Analysis of Sample of 441 AIs from CDBG/HOME Grantees: 

To estimate the percentage of grantees with outdated and current AIs, 
the sample data were weighted to make them representative of the 
population of grantees from which the sample is drawn. Our sample is 
stratified by region (10 HUD regions) and grant size (less than 
$500,000 and $500,000 or more in fiscal year 2009), with equal numbers 
of grantees being selected from each of 10 HUD regions.[Footnote 35] 
Since in our sample the probability of a grantee being selected varied 
by stratum, we assigned different weights, or sampling weights, to 
grantees in different strata when estimating population statistics 
(percentages) for the combined groups.[Footnote 36] 

To assess the precision of our estimates, we calculated 95 percent 
confidence intervals for each measure.[Footnote 37] Calculated from 
sample data, a confidence interval gives an estimated range of values 
that is likely to include the true measure of the population. For the 
estimated percentage of outdated AIs, we calculated a lower and upper 
bound at the 95 percent confidence level (there is a 95 percent 
probability that the actual percentage falls within the lower and 
upper bounds) of grantees by HUD region and by grant size category 
using raw data and the appropriate sampling weights.[Footnote 38] We 
used the standard errors of the estimates to calculate whether any 
differences between the grantees by region and grant size were 
statistically significant at the 95 percent confidence level. 

To evaluate the timeliness of AIs, we relied on HUD criteria, 
including the 1996, 2000, and 2004 guidance that recommended that they 
be updated every 3 to 5 years, and annually as necessary, and the 
findings of a 2009 internal HUD study on AI compliance, which 
concluded that AIs completed in the 1990s are outdated. Specifically, 
using a data collection instrument (DCI), we systematically noted the 
publication dates of all 441 AIs and also noted if no date was 
mentioned in the AI. We also collected information on the author's 
name, if available, and the number of pages. Based on this, we 
categorized 64 percent AIs as current. 

Analysis of Sample of 60 Current AIs: 

Further, to gain more information on the contents of AIs categorized 
as current (AIs completed from 2005 through 2010), we reviewed a 
nonrepresentative subset of 60 current AIs to determine the extent to 
which they contained sections that the HUD guidance suggests to 
include in AIs: (1) executive summary/introduction, (2) grantee's 
background data, (3) current fair housing legal status, (4) identified 
impediments, (5) recommendations, (6) time frames, and (7) signatures 
by chief elected officials. We chose the subset on a weighted 
geographic basis to ensure that it was reflective of the geographic 
distribution of the current AIs and reflected grantees in terms of the 
distribution of fiscal year 2009 grant sizes within region. We 
selected between five and seven grantees per HUD region weighted 
slightly more toward grantees with larger grants because they were 
more prevalent among the grantees with current AIs. As such, per 
region, we selected between one and three grantees with grant sizes 
less than $500,000 in fiscal year 2009 and between three and four 
grantees with grant sizes of $500,000 or more. While we took steps to 
help ensure that the 60 current AIs were reflective of the diversity 
in content in such documents by basing the selection on such factors 
as the grantees' geographic location and grant size, they are not 
representative of either all current AIs in our sample or of all 
current AIs generally. 

Analysis of Sample of 30 Current AIs: 

Finally, we reviewed 30 of the nonrepresentative subset of 60 current 
AIs to identify the types of potential impediments to fair housing 
choice that are commonly identified in such documents and to provide 
specific examples of such impediments. We selected a subset of 30 of 
the 60 current AIs and restricted our analysis to impediments 
summarized only in one of three possible sections of the AI: 
introduction, executive summary, or conclusion. To generate the sample 
of 30 current AIs from the larger subset of 60, we selected 3 current 
AIs from each region, 1 from small grantees (fiscal year 2009 grant 
amounts of less than $500,000), and 2 from large grantees since this 
weighting is reflective both of the overall distribution of grantees 
and of those with current AIs in which large grantees are more 
prevalent. During the course of our analysis, 5 of the original 30 
were replaced with AIs from other similar grantees (based on regions 
and grant size) from the subset of 60 AIs because they did not list 
impediments in any of the three sections. While we took steps to 
ensure that these 30 AIs were reflective of the diversity of the 
population of such documents, they are not representative of current 
AIs in our sample or of all AIs. 

Qualitative analyses were conducted to identify and code impediments 
listed in the 30 AIs. One GAO analyst identified the impediments 
described in either the introduction, executive summary, or conclusion 
of the AI and coded them from 1 to 13 using the types of possible 
impediments to fair housing choice described in HUD's 1996 Fair 
Housing Planning Guide. A second analyst independently verified them 
by reviewing the codes assigned by the first reviewer and then either 
indicating agreement with the first reviewer's codes or assigning a 
different code for later discussion with the first reviewer. If 
disagreements occurred, the GAO analysts discussed their differences 
and came to an agreement. Each AI contained a list of multiple 
impediments that were usually coded into one category each. Sometimes, 
however, individual impediments were coded into more than one category 
or multiple impediments within one AI were coded into the same 
category. 

Then, we compared the findings of our analysis of AIs' timeliness and 
assessment of the contents of AIs categorized as current with the 
results of HUD's 2009 AI study. We interviewed HUD officials from both 
headquarters and 10 field offices to gather their views on grantee's 
compliance to the affirmatively further fair housing (AFFH) 
requirement. 

To address the second objective, we reviewed and analyzed HUD's 
policies, procedures, and guidance for overseeing and enforcing the 
AFFH requirement, particularly pertaining to AIs, as well as gathering 
information on staff resource levels for doing this. We gathered 
information from select field offices on how they interpret and 
implement existing AFFH regulations and guidance and conducted a 
limited review of annual reports that are required by grantees for 
submission to HUD. Additionally, we obtained and reviewed data from 7 
of the 10 field offices we contacted on the number of times CPD staff 
obtained and/or reviewed AIs during on-site grantee monitoring reviews 
in 2009. The other 3 field offices did not provide the data as 
requested. 

To assess the extent to which HUD's general processes are sufficient 
in their design and implementation to help ensure grantees' compliance 
with AFFH documentation requirements, we reviewed the 2009 internal 
HUD study on AI compliance and oversight, obtained a senior HUD 
official's public testimony on the issue, and interviewed HUD 
officials at the HUD headquarters and officials in 3 of 10 regional 
offices and 7 of 81 field offices. We selected offices in a way that 
emphasized geographic diversity and the representation of 
jurisdictions with a large number of grantees or at greater risk of 
noncompliance as measured by the estimated incidence of grantees 
having submitted outdated or no AIs. During these site visits, we 
conducted a file review of compliance documents for grantees that were 
on file and met with several officials to discuss current enforcement 
and oversight activities, as well as the potential limitations to 
enforcing and overseeing AFFH activities. 

Analysis of Sample of 30 CAPERS from Grantees with Outdated AIs: 

To assess the usefulness of required AFFH documents for supervising 
grantee compliance for objective 2, we reviewed other required AFFH 
documents including 30 Consolidated Annual Performance and Evaluation 
Reports (CAPER) that grantees are required to submit annually to lay 
out grantees' progress in meeting their objectives for the use of CDBG 
and other funds. The purpose of the limited CAPER review was to 
determine the extent to which the CAPER included information about the 
status of the AI document, including the AI date, if any, the depth of 
discussion regarding the AI in the CAPER, and other information 
related to the timeliness of the AI. Specifically, we randomly drew a 
nongeneralizable sample of 30 grantees from the subset of 71 outdated 
AIs completed prior to 2003 in the original sample of 473 grantees and 
contacted HUD to obtain the latest CAPER for these 30 grantees. 
[Footnote 39] The subset was chosen on a weighted geographic basis to 
reflect the 10 HUD regions and size of CDBG/HOME grants received in 
fiscal year 2009. Between 1 and 4 AIs were selected from each region 
with slightly more from grantees with large awards ($500,000 or more) 
since these were more prevalent in the original sample and in the 
subset of outdated AIs. GAO has previously reported on internal 
control standards, as well as HUD's oversight of the CDBG program and 
federal fair lending. We consulted these reports as necessary to draw 
conclusions about HUD's AFFH oversight program. 

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

[End of section] 

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

U.S. Department Of Housing And Urban Development: 
Assistant Secretary for Fair Housing and Equal Opportunity: 
Washington, DC 20410-2000: 

September 3, 2018: 

Orice Williams: 
Director, Financial Markets and Community Investments: 
U.S. General Accounting Office: 
Washington, DC 20548: 

Re: HUD Response to GAO Draft Audit Report No. 250496. 

Dear Ms. Williams, 

The U.S. Department of Housing and Urban Development appreciates the 
Government Accountability Office's (GAO's) attention to the 
Department's Affirmatively Furthering Fair Housing (AFFH) mission and 
welcomes this opportunity to comment on GAO's "Review of Grantee 
Compliance and HUD Enforcement of AFFH Requirements." 

The Department's statutory responsibility to affirmatively further 
fair housing is an important civil rights authority intended to 
address years of racial discrimination and the legacy of segregation. 
HUD is committed to providing effective training, assistance, 
oversight, and enforcement to ensure that all recipients carry out 
their responsibility to aftirmatively further fair housing. 

HUD's Renewed Commitment to AFFH: 

We have made significant changes to ensure that communities receiving 
HUD funds are affirmatively furthering fair housing. This year, 
through HUD's Fiscal Year 2010 Notice of Funding Availability and 
General Section (General Section), the Department strengthened and 
clarified the AFFH requirement for any grantee not specifically 
exempted. To meet this requirement, an applicant for HUD funding must 
address how its proposed activities will help overcome impediments to 
fair housing choice that are identified in the "Analysis of 
Impediments to Fair Housing Choice" (AI) produced by the state and/or 
local government(s) in which the program activities occur, the effects 
of those impediments, and any other impediments in housing or urban 
development activities that the applicant identifies. In addition, the 
applicant must address at least one of the following objectives: (1) 
help overcome any impediments to fair housing choice related to the 
assisted program or activity itself; (2) promote racially, ethnically, 
and socioeconomically diverse communities; or (3) promote housing-
related opportunities that overcome the effects of past discrimination 
because of race, color, national origin, religion, sex, disability, 
and familial status. 

Through the General Section, the Department also has made AFFH a 
"policy priority" within HUD discretionary funding programs. This will 
allow the Department to encourage grantees to undertake comprehensive 
and innovative strategies to affirmatively further fair housing. 
Specifically, HUD may award additional points to applicants for 
certain competitive grants that engage in activities that go beyond 
the minimum affirmatively furthering fair housing requirements and 
undertake activities such as participating in regional strategies to 
promote integration, decreasing the concentration of poverty, and 
promoting visitability for persons with disabilities. 

Increased Review and Technical Assistance for AFFH Compliance: 

Since the start of FY2010, FHEO regional and field offices have 
increased the level of review of recipients' AIs within their 
jurisdictions. Some offices are requesting and reviewing Ms as part of 
their assessment of the FY2010 Consolidated Plans, Annual Action Plans 
and Consolidated Annual Performance Evaluations Reports (CAPERs). 
Others offices are selecting AIs of recipients with a higher risk 
under our internal risk assessment protocol. This fiscal year, the 
Department has reviewed the AIs of more than 300 recipient 
jurisdictions. 

Similar to GAO, we have found that many of these AIs are out of date 
or insufficient. We are educating recipients on how to comply with 
their affirmatively furthering fair housing responsibilities. Since 
January of this year, the Department has increased its training of 
recipients concerning compliance with the requirement to affirmatively 
further fair housing. The training covers the AI and the importance of 
its completion; what information should be included in the AI, 
including race and ethnicity data and data regarding other protected 
classes; who should participate in creating the AI; the consequences 
of not completing the AI; and how to report fair housing activities. 

In addition to group training, the Department has provided technical 
assistance to all types of jurisdictions that receive HUD funding, 
including cities, counties, and states. There are times when education 
and technical assistance are not enough and the Department has used 
enforcement and other actions to bring recipients into compliance with 
their affirmatively furthering fair housing responsibilities. In this 
regard, in accordance with program requirements, the Department also 
has required recipients to provide further assurances, including a 
commitment to provide an updated, more comprehensive AI, in order for 
HUD to accept their certification that they will affirmatively further 
fair housing. Certifying to affirmatively further fair housing is a 
requirement for the receipt of certain HUD funds. 

AI monitoring also is incorporated into HUD monitoring handbooks and 
checklists used for grantee compliance monitoring by the Office of 
Community Planning and Development (CPD). See, for example, the 
Community Planning and Development Monitoring Handbook, HUD Handbook 
6509.2 REV-6, Chapter 22, and the AI monitoring checklists contained 
in Community Planning and Development Monitoring Handbook Exhibits 22-
1, 22-2, and 22-3. 

Improving HUD's Capacity for Monitoring and Enforcing AFFH Compliance: 

As detailed in the GAO Report, we are striving to better use our 
existing resources to promote and enhance compliance with the AFFH 
requirement. We have identified areas where we can target activities 
in order to maximize the effect of our civil rights statutes. We are 
exploring where we can find greater efficiencies and reduce staff time 
on routine or administrative matters so that more time can be 
dedicated to affirmatively furthering fair housing. 

To strengthen staff skills, we are making investments in the capacity 
of staff in relation to affirmatively furthering fair housing. At 
HUD's National Fair Housing Training Academy, we are designing 
training for staff on reviewing submissions to ensure consistent and 
valid review criteria. To assist in the review of Annual Action Plans 
and Consolidated Plans for compliance with the AFFH certifications, we 
are developing uniform standards of review for use by all FHEO staff. 
This will streamline the assessment of recipients and indicate the 
appropriate remedy for any deficiency found. 

We will also enhance compliance reviews of recipients of HUD funding 
under various civil rights authorities, including Title VI and Section 
504 of the Rehabilitation Act of 1973, to include considerations of 
whether those recipients are furthering fair housing. We will 
strategically use our authority to review a variety of recipients and 
maximize the effect of each review. We believe these reviews will send 
a clear message to all funded recipients that HUD takes its and their 
AFFH obligations seriously. 

Again, we appreciate the opportunity to comment on the draft report. 
Should you or your staff have any questions or require additional 
information please contact Janet Hostetler at (202) 708-4252. 

Sincerely, 

Signed by: 

John Tranvina: 
Assistant Secretary: 

[End of section] 

Appendix III: GAO Contact and Staff Acknowledgments: 

GAO Contact: 

Orice Williams Brown, (202) 512-8678, or williamso@gao.gov: 

Staff Acknowledgments: 

In addition to the individual named above, Wesley Phillips, Assistant 
Director; Farah B. Angersola; Emily Chalmers; William Chatlos; 
Jennifer Cheung; Pamela Davidson; Laurie Ellington; Delores Hemsley; 
Simin Ho; Fred Jimenez; Reginald Jones; John McGrail; Mark Molino; and 
Dae Park made significant contributions to this report. 

[End of section] 

Footnotes: 

[1] Pub. L. No. 90-284, title VIII, 82 Stat. 81 (1968), as amended, 
codified at 42 U.S.C. §§ 3601-3619. 

[2] This review covers only the CDBG program, which provides block 
grants to states, cities, and counties for a variety of affordable 
housing activities, and the HOME program, which supports the 
production of affordable housing. Other HUD formula grant programs 
that must comply with AFFH requirements include the Housing 
Opportunities for Persons with AIDS (HOPWA) program, and the Emergency 
Shelter Grants (ESG) program. 

[3] U.S. Department of Housing and Urban Development, "HUD and Justice 
Department Announce Landmark Civil Rights Agreement in Westchester 
County" (news release, Aug. 10, 2009); DOJ, "Westchester County Agrees 
to Develop Hundreds of Units of Fair and Affordable Housing in 
Settlement of Federal Lawsuit" (news release, Aug. 10, 2009). 

[4] National Fair Housing Alliance, "Analysis of Impediments to Fair 
Housing Study" (1997); and "Still Separate and Unequal: The State of 
Fair Housing In America" (written testimony submitted by William R. 
Tisdale, Metropolitan Milwaukee Fair Housing Council, July 15, 2008). 

[5] See U.S. Department of Housing and Urban Development, Policy 
Development Division, Office of Policy Development and Research, 
"Analysis of Impediments Study" (Washington, D.C., 2009). According to 
HUD officials who conducted the study, the goal was to assess the 
extent to which the AIs were produced in accordance to HUD's 1996 Fair 
Housing Planning Guide. The study focused on a number of issues, 
including the timeliness of submissions of the AIs, and also 
categorized the impediments to fair housing choice as outlined in the 
AIs. 

[6] For example, John D. Transvina, Assistant Secretary for Fair 
Housing and Equal Opportunity, U.S. Department of Housing and Urban 
Development, statement before the Subcommittee on Housing and 
Community Opportunity, Committee on Financial Services, U.S. House of 
Representatives, Housing Fairness Act of 2009, Hearing on H.B. 476-- 
Jan. 20, 2010. 

[7] The initial sample of 473 was reduced to 466, after we determined 
that 7 grantees were not required to prepare AIs. For example, while 
HUD's fiscal year 2009 program list contained these 7 grantees, HUD 
officials said that, in fact, some of them were incorrectly labeled as 
program participants that year. 

[8] All national estimates for percentage of outdated AIs reported 
from the survey results have a margin of error within plus and minus 5 
percentage points at the 95 percent confidence level. 

[9] Both the 60 and 30 current AIs were selected to be reflective of 
the sample of 281 current AIs in terms of geographic diversity and 
grant size. 

[10] As discussed in this section, we reviewed 30 current AIs to 
identify commonly cited impediments in them. We also reviewed 60 
current AIs, including these 30 to determine their authorship and to 
assess their conformance with the suggested elements of the format in 
HUD's 1996 Fair Housing Planning Guide. 

[11] Section 104(b)(2) of the Housing and Community Development Act of 
1974, as amended (42 U.S.C. 5304) requires, among other things, that 
each CDBG grantee certify to HUD's satisfaction that the grantee will 
AFFH. Section 105 of the Cranston-Gonzalez National Affordable Housing 
Act, (42 U.S.C. 12705) established a requirement of a comprehensive 
housing strategy as a condition of a grantee receiving grants from 
HUD, primarily CDBG and HOME programs. The strategy includes a 
certification that grantees receiving the HUD grants will AFFH. 

[12] U.S. Department of Housing and Urban Development, Office of Fair 
Housing and Equal Opportunity, Fair Housing Planning Guide: Volume 1 
(Washington, D.C.: 1996). 

[13] U.S. Department of Housing and Urban Development, Policy 
Development Division, Office of Policy Development and Research, 
Analysis of Impediments Study (Washington, D.C.: 2009). 

[14] We defined grantees that received less than $500,000 as small and 
those receiving $500,000 or more as large. We found that grantees that 
received less than $500,000 of CDBG and HOME funds did not produce 
significantly more outdated AIs than those that received more funds. 
Grant size is a proxy for the size of the grantee, because funds are 
allotted based on population, among other factors. 

[15] All regional estimates for the percentage of outdated AIs 
reported from the survey results have a margin of error within plus 
and minus 15 percentage point at the 95 percent confidence level. 

[16] U.S. Department of Housing and Urban Development, Policy 
Development Division, Office of Policy Development and Research, 
Analysis of Impediments Study (Washington, D.C.: 2009). 

[17] These limitations include the use of unweighted population 
estimates. 

[18] Subprime lending generally involves the origination of mortgages 
to borrowers who may represent greater default risks than prime 
borrowers, due for example to lower credit scores, on terms that may 
increase the potential for default. For example, a common subprime 
mortgage product would be an adjustable rate mortgage in which the 
interest rate increases substantially after the expiration of a lower 
"teaser" rate. Some have alleged that subprime mortgage originators 
disproportionately targeted minority borrowers earlier in this decade, 
which resulted in disproportionate foreclosure rates amongst such 
borrowers in recent years. 

[19] As discussed in this report, HUD field office officials also 
identified at least one grantee that received funds without having 
prepared an AI. 

[20] We analyzed the length of all the AIs we received (current, 
outdated, and undated). 

[21] While this subset is not generalizable to the population of all 
grantees, we used several criteria in drawing it to help ensure that 
it reflected the characteristics of AIs that we received. For example, 
we chose the subset on a weighted geographic basis to help ensure that 
it reflected grantees located across HUD's 10 regional offices. We 
also sought to help ensure that the subset reflected jurisdictions' 
size as based on CDBG and HOME grant amounts in fiscal year 2009. 

[22] In selecting the 30 AIs to review, we included both large and 
small grantees from each of HUD's 10 regional offices. We identified 
all impediments and coded them from 1 to 13 based on the categories 
described in HUD's 1996 Fair Housing Planning Guide. A second reviewer 
then verified the coding. Appendix I contains more information on our 
methodology. 

[23] Other impediments that grantees cited included tenant selection 
procedures used by public housing agencies and other assisted/insured 
housing providers, sales of subsidized housing and possible 
displacement, property tax policies, planning and zoning boards, 
building codes, and fair housing enforcement. See chapter 5 of HUD's 
Fair Housing Planning Guide (1996). 

[24] Pub. L. No. 94-200, title III, 89 Stat. 1125, codified at 12 
U.S.C. §§ 2801-2810. HMDA requires lending institutions to collect and 
publicly disclose information about housing loans and applications for 
such loans, including the loan type and amount, property type, and 
income level and other characteristics of borrowers (such as 
ethnicity, race, and sex). All federally insured or regulated banks, 
credit unions, and savings associations with total assets exceeding 
$39 million, as of December 31, 2008, with a home or branch office in 
a metropolitan statistical area that originated any secured home 
purchase loans or refinancing are required to file HMDA data. 
Regulation C, 12 C.F.R. §§ 203.3(e)(1), 203.4 (2009); see also Home 
Mortgage Disclosure, 73 Fed. Reg. 78616 (Dec. 23, 2008) (establishing 
an adjustment from $37 million to $39 million). Further, most mortgage 
lending institutions located in a metropolitan statistical area must 
file HMDA data. 12 C.F.R. §§ 203.3(e)(2), 204.4 (2009). 

[25] For example, the Sarbanes-Oxley Act of 2002 requires senior 
officials of public companies to sign a certification that the 
companies financial statements are correct. 

[26] Fair housing testing may involve the use of individuals posing as 
renters or buyers to determine whether housing providers are complying 
with fair housing laws. 

[27] During our review, we interviewed officials from 3 HUD regional 
offices and 7 field offices. To facilitate the discussion in the 
report, we use the term "field office" to cover all 10 of the offices 
we contacted. 

[28] HUD received over $13 billion dollars in Recovery Act funding, 
including $1 billion to support projects within the Community 
Development Block Program which HUD staff must implement and oversee. 
See Pub. L. No. 111-5, American Recovery and Reinvestment Act of 2009, 
123 Stat. 115 (2009). 

[29] While this subset is not generalizable to all grantees with 
outdated AIs, we used criteria in drawing it to help ensure that they 
reflected the diversity seen in the sample of grantees with outdated 
AIs in our original sample. Specifically, we chose the subset on a 
weighted geographic basis to help ensure that it included grantees 
located across HUD's 10 regional offices. We also sought to ensure 
that the subset reflected grantees in terms of the size of CDBG grants 
they received in fiscal year 2009. We selected between one and four 
grantees per HUD region weighted slightly more toward grantees with 
larger grants since they were more prevalent grantees with AIs 
completed prior to 2005 in all but two HUD regions. 

[30] John D. Transvina, Assistant Secretary for Fair Housing and Equal 
Opportunity, U.S. Department of Housing and Urban Development, 
statement before the Subcommittee on Housing and Community 
Opportunity, Committee on Financial Services, U.S. House of 
Representatives, Housing Fairness Act of 2009, Hearing on H.B. 476--
Jan. 20, 2010. 

[31] U.S. Department of Housing and Urban Development, Community 
Development Block Grant, [hyperlink, 
http://www.comcon.org/programs/contact_cdbg.html], accessed in 
November 2009. 

[32] There were originally 1,209 CDBG and HOME grantees in the 
population. We chose a representative sample of 473 grantees from this 
population. However, HUD later informed us that 7 of the 473 grantees 
in our sample did not need to prepare an AI for various reasons (e.g., 
the jurisdiction was no longer a CDBG program participant). Because we 
can assume that the 7 out-of-scope grantees discovered in our sample 
represent an even larger number in the population, we estimated a new 
population count at the stratum level based on the number of out-of- 
scope grantees and the number of grantees sampled. Using this 
estimate, we adjusted the universe of CDBG and HOME grantees from 
1,209 to 1,190. 

[33] Excluded grantees were considered out-of-scope if they were 
either not CDBG program grantees or it was determined that they did 
not need to produce an AI. 

[34] Based on the number of jurisdictions in our starting sample, the 
raw response rate was 93 percent. We later determined that seven 
jurisdictions did not need to prepare an AI for various reasons, 
including that they were no longer participants in the CDBG program. 
Eliminating these jurisdictions reduced the sample size from 473 to 
466 and resulted in an adjusted response rate of 95 percent. 

[35] We independently selected a random sample of 48 grantees from 
each of 10 HUD regions (with the exception of Region VII where there 
were only 41 grantees in population). 

[36] We calculated the weights as: 

Nh 

wh = nh 

where, wh denotes the weight for the stratum (h=1, 2, …,20); Nh 
denotes the population for the hth stratum; and nh denotes the total 
number of survey responses for the hth stratum. We calculated the 
ratio estimate of the overall population as: 

R = (Shwh Si yhi)/(Shwh Si xhi) 

where, wh denotes the sample weight for the hth stratum; yhi 
represents the ith response of the variable y response the in the hth 
stratum; xhi represents the ith response of the variable x in the hth 
stratum; and R denotes a population estimate of the ratio. 

[37] Since our estimates were calculated from a stratified sample, we 
obtained the 95 percent confidence intervals of our population 
estimates by using methods detailed in: W.G. Cochran, Sampling 
Techniques, 3rd ed., Wiley Series in Probability and Mathematical 
Statistics, section 5.4 (New York, N.Y.: John Wiley & Sons, 1977), 95- 
96 and in: M.H. Hansen, W.N. Hurwitz, and W.G. Madow, Sample Survey 
Methods and Theory, Methods and Applications, Wiley Publications in 
Statistics, section 5.4 (New York, N.Y.: John Wiley & Sons, Inc., 
1953), 1: 189-194. 

[38] All regional estimates for percentage of outdated AIs reported 
from the survey results have a margin of error within plus and minus 
15 percentage points at the 95 percent confidence level. 

[39] By using 2003 instead of 2005 as cutoff, we are applying a more 
stringent definition of "outdated" than previously as a way of 
targeting grantees that likely relied on demographic data that were 20 
years old (e.g., 1990 census data). 

[End of section] 

GAO's Mission: 

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

Obtaining Copies of GAO Reports and Testimony: 

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

Order by Phone: 

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

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

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

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

Contact: 

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

Congressional Relations: 

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

Public Affairs: 

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