Fair Housing: HUD Needs Better Assurance That Intake and	 
Investigation Processes Are Consistently Thorough (31-OCT-05,	 
GAO-06-79).							 
                                                                 
Each year, the Department of Housing and Urban Development's	 
(HUD) Office of Fair Housing and Equal Opportunity (FHEO) and	 
related state and local Fair Housing Assistance Program (FHAP)	 
agencies receive and investigate several thousand complaints of  
housing discrimination. These activities, including required	 
conciliation attempts, are directed by HUD's standards, which are
based on law, regulation, and best practices. GAO's 2004 report  
examining trends in case outcomes raised questions about the	 
quality and consistency of the intake (the receipt of initial	 
inquiries) and investigation processes. This follow-up report	 
assesses the thoroughness of fair housing intake and		 
investigation (including conciliation) processes, and complainant
satisfaction with the process.					 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-79						        
    ACCNO:   A40679						        
  TITLE:     Fair Housing: HUD Needs Better Assurance That Intake and 
Investigation Processes Are Consistently Thorough		 
     DATE:   10/31/2005 
  SUBJECT:   Claims processing					 
	     Customer service					 
	     Discrimination					 
	     Grievance procedures				 
	     Housing						 
	     Internal controls					 
	     Investigations by federal agencies 		 
	     Noncompliance					 
	     Records management 				 
	     Standards						 
	     Surveys						 
	     Timeliness 					 
	     HUD Fair Housing Assistance Program		 
	     HUD Title Eight Automated Paperless		 
	     Tracking System					 
                                                                 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-79

Report to Congressional Requesters

October 2005

FAIR HOUSING

HUD Needs Better Assurance That Intake and Investigation Processes Are
Consistently Thorough

Contents

Tables

Figures

October 31, 2005Letter

The Honorable Paul S. Sarbanes Ranking Minority Member Committee on
Banking, Housing, and Urban Affairs United States Senate

The Honorable Jack Reed Ranking Minority Member Subcommittee on Housing
and Transportation Committee on Banking, Housing, and Urban Affairs United
States Senate

In 2004, the Department of Housing and Urban Development (HUD) and the 100
state and local agencies it has certified through its Fair Housing
Assistance Program (FHAP agencies) handled over 9,000 complaints  of 
housing discrimination under the Fair Housing Act (Title VIII of the Civil
Rights Act of 1968, as amended).1 The act generally prohibits
discrimination against minorities, persons with handicaps, and other
protected groups in the sale and rental of residential dwellings. The act
as amended provides HUD with enforcement powers and establishes a 100-day
benchmark for completing investigations of complaints.2 HUD's Office of
Fair Housing and Equal Opportunity (FHEO) administers the program and
handles complaints of housing discrimination. FHAP agencies also handle
complaints that allege violations of substantially similar state and local
laws.

To meet the Fair Housing Act's (Act) requirements, FHEO has developed a
process for receiving, investigating, and resolving housing discrimination
complaints. This process has three stages:

o intake, during which FHEO offices and FHAP agencies receive inquiries
from individuals (complainants), determine whether the inquiries involve a
potential violation of the Fair Housing Act (or equivalent state law), and
file fair housing complaints for those that do;3

o investigation, during which FHEO or FHAP agency investigators collect
evidence to determine whether reasonable cause exists to believe that a
discriminatory housing practice occurred or is about to occur and
simultaneously work with parties to conciliate, or reach a mutually
acceptable solution; and

o adjudication, during which an administrative law judge, another
administrative entity, or a federal or state court actually determines
whether a violation of the Act has occurred.

The Act requires that efforts be made to conciliate complaints-that is, to
reach a resolution acceptable to all parties-throughout the complaint
process, beginning with the intake stage. FHEO and FHAP agencies use an
automated case tracking system, the Title Eight Automated Paperless Office
Tracking System (TEAPOTS) to record information about complaints and key
steps in the investigative process.

HUD reported in 2002, that the incidence of consistent adverse treatment
against minority home seekers had declined over the last decade.
Nonetheless, a series of reports using paired testing to measure the level
of discrimination in the U. S. housing market found that about 20 percent
of the time a member of a protected group will experience discrimination
when attempting to rent an apartment.

This report, which focuses on intake, investigation, and conciliation (the
complaint process), is the second in a series on HUD's handling of housing
discrimination complaints. GAO's earlier report on fair housing found,
among other things, that as of September 2003 more than a third (39
percent) of open housing discrimination investigations had passed the
100-day benchmark.4 GAO noted at the time that FHEO and the FHAP agencies
were trying to speed their efforts to resolve these cases. Further, GAO
reported that FHEO hub directors cited tension between the need to meet
the 100-day benchmark and the need to conduct a thorough investigation,
stating that at times one goal cannot be achieved without some cost to the
other. The report also found that increasingly the most common outcome of
housing discrimination investigations was a finding of "no reasonable
cause" to indicate discrimination occurred. In 2004, almost half of all
investigations closed had a finding of no-cause and less than 7 percent
had a finding of reasonable cause. About one third were conciliated.

This report addresses the thoroughness of the process for resolving
housing discrimination complaints-that is, the extent to which FHEO and
FHAP agencies meet HUD standards, policies, and best practices in a timely
manner. GAO reviewed

o the thoroughness and timeliness of FHEO and FHAP agencies' efforts
during the intake process;

o the extent to which FHEO and FHAP agencies ensure the thoroughness of
investigations, regardless of the outcome;

o the extent to which FHEO and FHAP agencies attempt to conciliate fair
housing complaints; and

o complainants' satisfaction with the thoroughness, fairness, and outcomes
of the fair housing process.

To assess the thoroughness of the intake process, we asked FHEO offices
and FHAP agencies around the country to keep logs of inquiries they
received during a 4-week period and measured (1) the extent to which these
inquiries were in FHEO's information system and resulted in fair housing
complaints and (2) the timeliness of the responses. The centers that
participated in the log exercise represented 78  percent of investigations
in 2004. We also made "test" telephone calls to 46 selected intake centers
(one FHEO office in each of HUD's 10 regions and all 36 statewide FHAP
agencies) during which GAO analysts posing as complainants contacted
intake staff to file a mock complaint. We used these calls to measure the
time it took to ultimately reach a staff person or to receive a return
call. We also determined the extent to which, during the initial contact,
intake staff gathered necessary information for deciding whether to
undertake a full investigation, including whether Title VIII appeared to
cover the alleged incident. Because of the limitations of this sample, the
results are not generalizable to all potential complainants of housing
discrimination. To assess the thoroughness of investigations and efforts
to conciliate complaints, we reviewed the documentation in 197 randomly
selected case files of housing discrimination cases from around the
country that were completed during the second half of calendar year 2004
with outcomes of no-cause or that were closed administratively or
conciliated. We used a structured data collection instrument to record
whether the files contained specific evidence of investigation steps and
their timing. We sampled a sufficient number of case files to permit
national estimates-that is, to permit statistical generalization to all
cases closed with these outcomes during the second half of 2004. Using the
same data collection instrument, we also reviewed 12 of the 15 FHEO cases
with reasonable cause outcomes that completed the adjudication process
during this period.5 In addition, for each case we compared selected file
contents with TEAPOTS records used to track the case. We did not assess
whether investigations reached an appropriate decision regarding any
specific fair housing inquiry or investigation. To measure complainants'
satisfaction with the complaint process, we conducted a nationwide
telephone survey of a random sample of 575 complainants in housing
discrimination cases that were investigated and closed between July 2004
and December 2004. The survey had an overall 38 percent response rate.6
Appendix I contains a detailed description of our methodology. In
addition, our detailed survey results can be found on GAO's Web site at
www.gao.gov. We conducted our work from September 2004 to October 2005 in
Atlanta; Chicago; San Francisco and Oakland; Kansas City, Kansas;
Baltimore; Columbia, South Carolina; and Washington, D.C., in accordance
with generally accepted government auditing standards.

Results in Brief

While often responding promptly and providing useful guidance and
information, FHEO and FHAP agencies were not always thorough or timely in
carrying out intake activities; moreover, missing or inconsistent data
suggests that TEAPOTS may be of limited usefulness as a management control
over the intake process. First, we found that potential complainants may
encounter difficulty in making initial contact with a FHEO or FHAP agency
staff person. For example, although fair housing agencies receive
allegations of discrimination principally via telephone, in five locations
we called, the agency did not return our test calls, even after three
attempts. Similarly, about 30 percent of complainants we surveyed-while
generally expressing satisfaction with aspects of the intake
phase-reported that it was somewhat or very difficult to reach someone the
first time they contacted a fair housing agency. Second, the agencies did
not consistently obtain information in a way that was expeditious for
complainants or that met recommended guidelines, as can be seen in the
following examples:

o The intake centers to which we placed test calls did not consistently
collect information that would help them recontact the complainant or
assess the urgency of the situation.

o Among our test calls, none of the intake centers collected all of the
information during the initial contact considered by HUD policy to be
critical to collect at intake, including the name and telephone number of
the respondent (the person or persons who allegedly committed the
prohibited discriminatory act). Rather, most of the agencies we contacted
used as their primary means of collecting information a written intake
form, which must be mailed to the complainant, filled out, and mailed back
to the agency, adding to the time required to act on the inquiry.

o Of the 306 inquiries that the centers, during the period of our review,
found met HUD's initial criteria for filing a complaint, only half of the
time did FHEO and FHAP agencies complete this process within the 20-day
benchmark period.

Our comparison of logged contacts with TEAPOTS data indicates that the
system's usefulness as a management control may be limited. While the
intake centers logged 2,000 unique new potential Title VIII violations,
only 631 (32 percent) were entered into TEAPOTS during the period of our
review.7 Further, because the initial contact dates shown in the logs were
sometimes earlier than the corresponding date of first contact shown in
TEAPOTS, HUD's use of TEAPOTS data overstates its performance in meeting
the 20-day intake timeliness benchmark. About 40 percent of inquiries that
completed the intake process took more than 20 days to complete.

Evidence demonstrating the thoroughness of investigations-that is, the
extent to which investigative standards were met, recommended procedures
were followed or were timely, or internal control measures were used and
documented-varied among the 197 case files and corresponding TEAPOTS
records. First, some files lacked documentation showing that investigators
had carried out certain required actions, such as sending complaint
notices, or copies of amended complaints, to both complainants and
respondents; sending copies of final closure notices; preparing final
investigative reports; and obtaining supervisory approval of final
investigative reports and letters describing the agency's final
determination. Second, an estimated 62 percent of the complaint files did
not contain detailed investigative plans that HUD guidance strongly
encourages.8 Further, while HUD has recommended procedures for
interviewing complainants and respondents or making on-site visits, an
estimated 28 percent of the files did not include evidence of interviews
with respondents, while an estimated 73 percent showed no evidence of
on-site visits. Third, we found that while HUD relies on TEAPOTS as a
control to assure that investigations meet statutory and regulatory
requirements, some FHEO and FHAP agencies did not include detailed
information about cases that would permit effective monitoring in the
system. Among complainants we surveyed, about one-third stated that they
believed their investigations had not been very thorough or at all
thorough. FHEO and FHAP agency officials and staff said that resource
shortages often presented a challenge in meeting program guidelines and
noted that investigative and management review steps that actually
occurred may not have been documented.

Complainants whose cases were closed through conciliation were the most
satisfied with their case outcomes; however, our review of case files and
corresponding TEAPOTS records found inconsistent documentation of
conciliation attempts. HUD requires FHEO and FHAP agencies to document
conciliation attempts with each party and record in writing the terms and
conditions of any agreement. The written agreements must be signed by all
parties and approved by HUD or the FHAP agency. Our review of the case
files and TEAPOTS records disclosed the following:

o In an estimated 36 percent of the cases, the files contained no evidence
that complainants had been contacted to attempt conciliation, and an
estimated 32 percent of the files contained no evidence that respondents
had been contacted.

o Among those cases that HUD closed with an outcome of conciliation, an
estimated 9 percent did not have written agreements in the case file.9
Among those that did contain written agreements, about 5 percent did not
contain signatures of all parties, and about 10 percent showed no evidence
of approval by HUD or the FHAP agency.10

o Information in TEAPOTS regarding conciliation efforts varied widely,
from a citation of "conciliation discussed" to the inclusion of
significant details.

Some surveyed complainants reported experiencing at least some pressure to
conciliate their complaints, most commonly because they feared losing
their cases. Others reported that they were not offered help in resolving
their complaints. HUD officials said that the statutory mandate inherently
creates some pressure to conciliate complaints and that such pressure is
not attributable to HUD's administration of the statute.

While survey respondents reported a number of positive impressions,
particularly with the intake phase of their complaints, they generally
expressed more dissatisfaction than satisfaction with both the overall
process and the case outcome. About 71 percent of complainants were
somewhat or very satisfied with the intake process, citing clarity of
information about steps in the process and generally professional and
courteous treatment. Nevertheless, we estimate that about half of all
complainants in cases closed during our survey period were either somewhat
or very dissatisfied with the fair housing complaint process overall.
Similarly, nearly 60 percent were dissatisfied with the outcome of the
process, and almost 40 percent would be unlikely to file a complaint in
the future. Complainants' satisfaction varied among stages of the
complaint process and outcomes.

o Some complainants viewed the intake process favorably. Although
complainants rated the investigation stage less favorable than intake,
over half were satisfied with the investigation stage. However, a
substantial number rated poorly certain aspects of intake and
investigation.

o Complainants whose cases ended with a finding of no-cause were the most
dissatisfied with the complaint process (72 percent dissatisfied, 23
percent satisfied) and with specific aspects of investigations.
Complainants whose cases were conciliated were the most satisfied (75
percent versus 25 percent, respectively).

Complainants' satisfaction was not linked to the type of agency conducting
the investigation (FHEO or FHAP).

We are recommending  that HUD consider a number of strategies to help
assure that the fair housing complaint process is thorough and timely,
including steps to facilitate complainants' initial inquiries, improve
data needed to monitor timeliness of intake activities, improve planning
and documentation of investigations, and help ensure that conciliation is
available to all complaints.

HUD's General Deputy Assistant Secretary for Fair Housing and Equal
Opportunity provided written comments on a draft of this report indicating
general agreement with our findings and recommendations. The comments also
included technical clarifications, which we have incorporated into this
report as appropriate. The General Deputy Assistant Secretary's letter is
reprinted in appendix II.

Background

The Fair Housing Act is the most comprehensive of the federal statutes
that prohibit discrimination in the rental and sale of housing.11 Passed
in 1968 and amended in 1988, the Act prohibits discrimination on the basis
of color, family status, handicap, national origin, race, religion, and
sex.12 It applies to a number of what are termed "issues," including
discrimination in the sale, rental, advertising, and financing of housing;
in the provision of brokerage services; and in other activities related to
residential real estate transactions.13 Generally, the Act covers all
dwellings-that is, buildings designed to be used wholly or in part as
residences and land where a dwelling will be located.14 When first enacted
in 1968, the Fair Housing Act's administrative enforcement process was
limited principally to conciliation. In 1988, Congress strengthened HUD's
authority and established a comprehensive administrative process to
enforce the law, but conciliation remained a primary feature.15

The Act gives HUD, private persons, and the U.S. Attorney General tools
and remedies to enforce the antidiscrimination provisions. Using HUD's
administrative process, individuals who believe they have experienced
discrimination in a housing-related situation can file a complaint that
HUD may then investigate and resolve. Individuals may also elect to file
suit in civil court rather than using the administrative procedure set out
in the act. The Attorney General can bring a civil action in cases that
show a pattern of discriminatory practices.

The Act Is Administered by FHEO and FHAP Agencies

FHEO has staff in each of HUD's 10 regional offices, or hubs, who respond
to complaints (see fig. 1). Agencies certified to participate in HUD's
Fair Housing Assistance Program (FHAP) and receive funding from HUD for
handling fair housing complaints are obligated to comply with FHEO's
reporting and record maintenance requirements, must agree to on-site
technical assistance provided by HUD, and are obligated to implement
certain policies and procedures. FHAP agencies must be in states or
localities whose laws provide rights and remedies that are substantially
similar to those in the Act-for example, local laws must provide for the
same 100-day benchmark for investigations that is stipulated in the Act.
FHEO offices refer complaints alleging violations of state and local fair
housing laws to FHAP agencies-for example, a certified state office of
civil rights. Currently, there are 100 of these agencies around the
country.

Figure 1: FHAP Agencies within HUD's 10 Regions

FHEO staff has responsibility for the intake, investigation, and
resolution of some of these complaints. Aggrieved persons may also go
directly to FHAP agencies, which then perform the intake process. If an
aggrieved party contacts a FHEO office regarding discrimination that
allegedly occurred in a state or locality that has a FHEO-certified
"substantially equivalent" state or local agency (that is, a FHAP agency),
FHEO will complete the intake process and refer the complaint to that
agency for enforcement.

In 2004, FHEO and the FHAP agencies received approximately 9,500 filed
complaints (see fig. 2). For this same period, only 5 percent of the
closed case files resulted in reasonable cause outcomes.16 HUD reimburses
FHAP agencies for carrying out investigations once FHEO has reviewed the
completed cases. Along with reviewing cases to determine whether HUD
should pay for services rendered, FHEO monitors FHAP agencies and provides
technical assistance.17 FHEO monitors the fair housing enforcement efforts
through TEAPOTS.

Our last report identified a number of human capital challenges facing
FHEO, including the number and skill level of FHEO staff, the quality and
effectiveness of training, and other issues. An FHEO official noted that
the staff shortage affected not only enforcement of the Act, but also
FHEO's other responsibilities, forcing managers to assume heavier
caseloads and professional staff to perform administrative duties rather
than concentrating on the complaint process. The total number of full-time
equivalents (FTE)18 in FHEO has fluctuated over the last 10 years, falling
from a high of 750 in fiscal year 1994 to a low of 579 in fiscal year
2000. In fiscal year 2004, FHEO had 650 FTEs.

Figure 2: Number of Complaints Filed in Each FHEO Region during Calendar
Year 2004, FHAP vs. HUD

Complaint Process Includes Intake and Investigation Phases

The complaint process beginning at intake represents the initial contact a
complainant has with an agency responsible for enforcing the Act or
equivalent state law. Figure 3 describes the complaint process for
HUD-investigated complaints. FHAP agencies would follow a similar process.
In the intake stage, FHEO hubs and FHAP agencies receive inquiries by
telephone, fax, mail, in person, or over the Internet. Intake staff record
inquiries in TEAPOTS, interview complainants, and may do other
research-for example, searches of public records-to see if enough
information exists to support filing a formal complaint. This process is
known as "perfecting" a complaint. In order to be perfected, a complaint
must

o contain the required four elements of a Title VIII complaint: the names
and addresses of the person alleging the discriminatory practice and the
respondent, a description and the address of the dwelling involved, and a
statement of the facts leading to the allegation; and

o satisfy the Act's jurisdictional requirements that the complainant has
standing to file the complaint; that the respondent, dwelling, subject
matter of discrimination (e.g., refusal to rent or sell) and the basis
(e.g., race, color, or familial status) for the alleged discrimination are
covered by the Act; and that the complaint has been filed within a year of
the last occurrence of the alleged discriminatory practice.

Hub directors decide which complaints meet these criteria and become
perfected complaints. Complaints that do not meet the criteria are
dismissed. Intake staff record information about perfected complaints in
TEAPOTS, have complainants sign the complaints, send letters notifying
complainants and respondents about the complaint and the process that will
be used to address it, and send the complaint file to an investigator.19
FHEO's Title VIII Intake, Investigation, and Conciliation Handbook
(Handbook) sets a 20-day benchmark for completing the intake stage for
these cases, but a 5-day benchmark for cases that it first takes in and
then refers to FHAP agencies.

Complaints that are perfected proceed to an investigation. During this
stage, FHEO and FHAP agencies gather evidence to determine whether a
violation of the Act or a state or local housing law has occurred or is
about to occur. The Handbook provides guidance for investigators but notes
that investigations may vary. Agency guidance directs that directors of
FHEO's hub offices review the results of completed investigations to
determine whether reasonable cause exists to believe that a discriminatory
housing practice has taken place or could take place. With the concurrence
of the relevant HUD regional counsel, the hub director issues a
determination and directs the regional counsel to issue a charge, or short
written statement of facts, that led to the decision. In a March 6, 2003
memorandum, HUD's Office of General Counsel (OGC) in headquarters
requested that regional counsels send OGC's Office of Fair Housing the
final draft of any charge that they propose to file and that they not file
charges until they have received a response from OGC's Office of Fair
Housing. Figure 3 provides an overview of HUD's basic fair housing
complaint process, including timeliness benchmarks established by the Act
or agency guidance.

Figure 3: HUD's Fair Housing Complaint Process

An investigation can be closed at any point for administrative reasons or
through conciliation. Cases are closed administratively for several
reasons-for instance, when a complainant withdraws from the case or cannot
be located.  The Act requires HUD to make conciliation efforts throughout
the complaint process, beginning when the complaint is filed and
continuing until the charge is filed or the case dismissed. The Handbook
and federal regulations allow investigators to make conciliation efforts,
but the regulations also state that generally officers, employees, and
agents not associated with the case will attempt conciliation.
Conciliation agreements are intended to protect the public interest
through provisions such as requiring respondents to file periodic reports
with HUD. When a conciliation agreement is reached, the Act authorizes the
Department of Justice to enforce the agreement in the event of a breach.
In 2004, FHEO hubs and FHAP agencies closed about one-third of their cases
via conciliation.

The Act set a deadline of 100 days from the date the complaint is filed
for completing an investigation or conciliating or otherwise closing a
case, unless doing so is "impracticable." If the investigation cannot be
completed within this time frame, FHEO or the FHAP agency must notify the
complainant and respondent in writing in what is called the "100-day
letter."20 In our previous report, we found that the number of
investigations completed within 100 days by the FHEO or FHAP agencies
increased significantly after 2001, partly in response to FHEO's
initiative to reduce aged cases.

Test Calls and Analysis of Log Data Raise Questions about Thoroughness of
Intake Process and Effectiveness of Controls

Although often responding promptly and providing useful guidance and
information, our test calls and analysis of contact logs found that FHEO
and FHAP agencies were not always thorough or timely in carrying out
intake activities. Our test calls, while not generalizable, suggest that
potential complainants may have difficulty in making initial contact with
an intake staff person; moreover, 30 percent of the complainants we
surveyed reported such difficulty. Our test calls also showed that FHEO
and FHAP agency staff sometimes did not seek information needed to
determine whether a potential violation of the Act had taken place and to
file a formal complaint, or gather limited information that might help the
agency recontact the complainant or assess the urgency of the situation.
Among the logged contacts that the agencies determined were potential
violations of the Act and were recorded in TEAPOTS, half resulted in
formal complaints. However, only 57 percent of these completed the process
within the 20-day benchmark. Additionally, missing or inconsistent data
suggests that TEAPOTS may be of limited usefulness as a management control
over the intake process.

Agency Staff Did Not Always Return Test Calls or Collect Initial Intake
Information Promptly

The intake process is the first contact prospective complainants have with
the agencies responsible for enforcing the Act or an equivalent state law.
Depending on the quality of intake, potential complainants may or may not
feel comfortable continuing the process, and those who do not may give up
on pursuing their complaints. Thus, the agency's initial response to
complainants plays an important role in the fair housing complaint
process. However, our test calls revealed some potentially serious lapses
in agencies' responses to complainants' inquiries.

First, we found that agencies did not always respond promptly to initial
attempts to contact them to file a complaint and that because of
requirements that some agencies imposed, trying to file a complaint could
be a challenging process. In 5 of the 46 calls, the agency did not return
the test call, even after 3 attempts.21 In another 2 cases, the intake
organization required that the caller provide intake information via the
Internet or in person. As shown in figure 4, in 20 of the remaining 39
test cases, the caller spoke with a live person on initial contact. Of the
9 calls requiring a callback, 6 were not returned within 1 business day,
and 3 were not returned for 3 or more days. Our survey of complainants
suggests that they experienced similar difficulties to ours in contacting
intake staff. An estimated 30 percent noted that it was either somewhat or
very difficult to reach a live person the first time they contacted a fair
housing agency, and 34 percent said they had difficulty contacting staff
after the initial contact. These percentages were relatively constant
regardless of whether FHEO or a FHAP agency handled the case or its
outcome, with one exception. Complainants whose cases were conciliated
reported that they had less difficulty contacting staff than complainants
whose cases were closed with other outcomes.

Figure 4: Selected Results of Test Calls

We also found that intake staff did not seem to display a sense of urgency
in dealing with complaints. Over half of the agencies (23 of 39) relied
primarily on a form that the complainant must fill out (HUD-903 or state
equivalent) to collect the information needed to begin an investigation,
and in the initial phone call requested little more than the complainant's
name and mailing address. Using such a form to gather information for a
potential complaint could take a week or more-during which the caller
could lose a housing opportunity. Two other agencies would not mail a
complaint form, insisting that the caller come in to the office to file a
complaint. However, information from contact logs that the FHAP agencies
and FHEO offices maintained for 4 weeks, at our request, showed that the
most prevalent mode of contact is telephone, and that walk-in and Internet
contacts represented less than 5 percent. Given this situation, requiring
potential complainants to appear in person added an additional challenge
that could potentially make it difficult for a complainant to continue
with the process. Further, a test caller to one of these agencies,
stressing the urgency of her situation, was informed that filing a
complaint was a "slow process" and that her complaint would not be acted
on for some time, whether intake was done over the phone or via the
organization's form. FHEO's annual performance goals do not include goals
for the time it takes to return initial contacts from complainants.
However, FHEO has established a 20-day benchmark for completing the intake
process, starting with the date that the initial inquiry is recorded in
TEAPOTS. In commenting on a draft of this report, HUD's General Deputy
Assistant Secretary for Fair Housing and Equal Opportunity stated that the
agency tracks the time it takes to file a complaint from the point of
initial contact, and that a new initiative, the FHEO-OGC Case Processing
Research Project, is expected to assist with decision making during the
intake process since it uses a triage system to determine case
complexity.22

Despite any inconveniences, when our test callers did reach the agencies,
the staff treated them well. In none of our test cases did hold time
exceed 3 minutes, and staff at several agencies spoke extensively with
test callers, answering questions and providing guidance and information
on the process. Our survey of former complainants that completed the
investigation process showed similar findings. While complainants had
difficulty reaching an agency, once they did, more than half said that
agency staff did either a good or an excellent job of explaining the
process and timing of each step.

FHEO and FHAP Agencies Collected Limited Information during Initial Intake
Contacts

When collecting intake information during our test calls, FHEO and FHAP
agency staff focused primarily on collecting the complainant's name,
address, and protected class, as well as a description of the
discriminatory act. Staff sometimes did not ask for other information that
would be helpful in recontacting the complainant or assessing the urgency
of the situation. To systematically assess the thoroughness of the intake
test calls, we identified criteria from the Act (the minimum elements of
information needed to proceed with the complaint), HUD's Title VIII
Handbook, and training materials from the National Fair Housing Training
Academy. Additionally, we obtained information on best practices from a
fair housing advocacy group as well as HUD's training materials and
interviews with agency officials. We categorized these criteria at four
levels:

o Level 1-information that, according to HUD policy, should always be
collected during intake, though not necessarily during the first contact,
regardless of the basis of the complaint or the protected class.

o Level 2-information that is potentially applicable to all complaints and
that should be collected during the intake process.

o Level 3-information that is relevant to a particular basis or protected
class-that is, information necessary to determine, for example, whether
the complainant met a certain protected class (e.g., handicap or familial
status).

o Level 4-information that is considered to be a best practice-for
example, information that may be used for testing.

We also discussed these criteria and the designated levels with agency
officials.

We measured the percentage of information elements collected at each of
these levels during our test calls. The elements associated with each
level, and the results we observed during our test calls, are shown in
figure 5.

While level 1 information should always be collected in order to proceed
with a complaint, some of this information may not always need to be
collected during the initial contact. However, HUD policy recommends that
staff obtain as much information as possible during the initial intake
interview. We also believe that level 1 items should, with little
exception, be collected as part of the initial contact in order to have
information necessary to recontact the complainant and determine the
urgency of the situation.

Figure 5: Percentage of Questions Asked, by Level

Note: Cases in which the test caller volunteered the information item are
included in the count of the number of cases in which intake staff asked
for the information. The total number of cases is 39.

On average, intake staff collected approximately 44 percent of level 1
information, which is identified by HUD policy as critical to collect
during intake.23 For example, 21 agencies did not ask for the respondent's
first and last name, and 15 did not ask for the respondent's organization.
As indicated in figure 5, in only 8 agency test cases did intake staff ask
the complainant for the name of an alternate contact person and in only 7
agency test cases did the intake staff request a work number-an important
piece of information because, according to FHEO officials, many
complainants are in the process of moving and are difficult to recontact.
Also, intake staff about half of the time asked for the respondent's
telephone number. Further, the agencies collected little information
beyond level 1. On average, they gathered 8 percent of level 2, 11 percent
of level 3, and 9 percent of level 4 information. Some level 4 (best
practice) information that was most often collected during our case study
includes the respondent's gender (13 test calls), respondent's race (9
test calls), and what the complainant hoped to see as a result of filing a
complaint (8 test calls). (See fig. 5.)

Although HUD policy recommends that intake staff obtain as much
information as possible regarding the aggrieved person's allegations
during the initial intake interview, the amount of information collected
on initial contact varied significantly by agency. One location collected
nothing beyond the complainant's name and mailing address, while another
collected up to 80 percent of the level 1 data. However, in none of the
test calls did intake staff collect 100 percent of the level 1
information. While most agencies appeared to collect the remainder of the
critical information through their intake forms (either the HUD-903 form
or state equivalent), this practice prevents the agency from taking any
further action on the complaint until a signed form is received.24
According to HUD officials, the revised Title VIII Handbook contains
standards for information that HUD will collect during initial contact
with a complainant. However, these and other standards may not apply to
FHAP agencies since their certification by HUD does not ensure that they
follow identical procedures.

The time it takes to receive the form can delay the enforcement process,
potentially resulting not only in the loss of a housing opportunity but
also in complainants becoming frustrated with the process and deciding not
to pursue their complaint. In particular, some complainants may need
urgent attention, such as when they are about to become homeless because
they are being evicted from their rental home or apartment, or are losing
the home they own in a foreclosure. HUD has authority under the Act in
such urgent circumstances to take prompt judicial action by authorizing
the Attorney General to initiate a civil action seeking appropriate
temporary or preliminary relief pending final disposition of the
complaint. HUD's Title VIII Handbook establishes that intake staff has a
critical role in identifying when a complaint may involve a situation
warranting prompt judicial action. Time is critical, and the efforts of
the intake staff are helpful in gathering sufficient information for
determining when prompt judicial action may be necessary. This is also
important for intake by the FHAP agencies, since they must have
substantially equivalent authority to seek prompt judicial action.

Intake Log Data Indicate That TEAPOTS Is of Limited Effectiveness as
Management Control of Intake Process

Our prior report noted that while FHEO offices kept records of potentially
Title VIII-related contacts, it had not required FHAP agencies to do so.
(For that reason, when preparing our prior report, we were unable to
determine the extent to which FHAP agencies met the goal of perfecting
complaints within 20 days of initial contacts.) FHAP agencies typically
entered information only for perfected complaints. Accordingly, we
recommended that HUD ensure that the automated case-tracking system
(TEAPOTS) include complete, reliable data on key dates in the intake stage
for FHAP agencies. The latest version of HUD's Title VIII Handbook (issued
in May 2005) requires FHEO intake staff to record in TEAPOTS each inquiry
that is potentially Title VIII-related, regardless of whether it results
in a perfected complaint.

Our comparison of information from the logs-that intake centers kept, at
our request, during February and March 2005-with TEAPOTS data highlights
the need for reliable information regarding potentially Title VIII-related
contacts and the dates of their occurrence. First, the analysis showed
that a substantial number of potentially Title VIII-related contacts (68
percent) were not entered in TEAPOTS, and of those that were, about half
resulted in perfected complaints. While there are valid reasons for this
"attrition," our results suggest that staff are not recording in TEAPOTS a
substantial number of potentially Title VIII-related contacts (inquiries
in which the caller alleges housing discrimination and intake staff
believe the call represents a potential Title VIII violation). Further, we
found that, for those contacts that were entered into TEAPOTS, the initial
contact dates shown in the logs were sometimes earlier than the
corresponding date of first contact shown in TEAPOTS. Thus, HUD's use of
TEAPOTS data is likely overstating its performance in meeting its 20-day
intake timeliness benchmark. Without assurance that TEAPOTS is being used
consistently, HUD is unable to account for potentially Title VIII-related
contacts that do not appear in the system or accurately measure timeliness
of those that are recorded in the TEAPOTS system, thus limiting TEAPOTS'
utility as a management control.

Many Logged Contacts Did Not Appear in TEAPOTS

To determine the volume of intake-related contacts received by fair
housing agencies and the proportion of these contacts that resulted in
perfected complaints, we had 47 sites record all incoming
contacts-telephone calls, walk-ins, and Internet queries-that were related
to fair housing over a 4-week period (February 22 through March 21,
2005).25 The 32 state FHAP agencies, 5 local FHAP agencies, and 10 FHEO
offices that participated in the log exercise represented 78  percent of
the volume of investigations in 2004. Specifically, we asked the sites to
record the date, method, and purpose of each contact; the name of the
person making the contact; whether the contact alleged having experienced
housing discrimination; and whether the intake staff agreed that the
matter pertained to a potentially Title VIII-related complaint.

During our tracking period, the sites recorded a total of 9,655 contacts.
As shown in figure 6, the majority (80 percent) of contacts for which we
had complete data was by telephone, and 42 percent were new potential
complaints. Furthermore, a sizable number of initial contacts
(approximately 24 percent) did not pertain to fair housing (confirming, as
our prior report noted, that intake analysts receive numerous contacts
that are not related to fair housing.) The time necessary for handling the
calls can place an additional burden on FHEO's limited resources.

Figure 6: Nature and Method of Contacts Received by Fair Housing Agencies

Note: The "nature of contact" pie chart does not include data for the
1,274 contacts from one agency, where staff did not record the nature of
contact. In addition, totals do not equal 100 percent due to rounding and
missing data.

Our review of the log also showed that a sizable number of new potential
complaints that intake staff believed could involve a Title VIII violation
did not result in perfected complaints in TEAPOTS (see fig. 7):

o Agency staff coded 2,000 contacts as coming from a named individual
whose allegations they believed both involved a new potential complaint
and pertained to a potentially Title VIII-related violation.26

o Of these 2,000 individuals, we were able to match 631 to a new inquiry
shown in TEAPOTS.

o Of the 631 inquiries, 306 are shown in TEAPOTS as perfected complaints.

Figure 7: Intake Process Attrition

We sorted these data into two groups: contacts recorded by FHEO offices
and contacts recorded by FHAP agencies. We found that the attrition rates
differed somewhat between these groups. At FHEO sites, intake staff
identified 1,347 unique individuals as having potentially valid new Title
VIII complaints. Of these, 506, or 38 percent, were shown as unique
inquiries in TEAPOTS, and 216 of those-16 percent of the original
1,347-resulted in perfected complaints shown in the automated system at
the time of our analysis. At the same time, FHAP sites identified 620
unique individuals27 with potential new Title VIII violations, of which
92, or 15 percent, were shown as unique inquiries in TEAPOTS and 66, or 11
percent, became perfected complaints.28

Attrition can occur for a number of reasons. For example, because some
state laws contain additional protected classes, some calls that FHEO
offices receive may be for matters that are covered under a FHAP agency's
jurisdiction. Also, intake staff may believe that a contact pertains to a
valid Title VIII violation at the outset but may later find out that the
respondent is exempt from Title VIII or that the 1-year statute of
limitations has expired.  Furthermore, the intake process is sometimes
terminated because a complainant either does not cooperate with agency
staff or resolves the issue with the respondent and voluntarily
discontinues the complaint process. Finally, it is possible that, during
our process of matching names in the log to TEAPOTS records, a small
number of matches were not made, either due to misspellings of names or
timings of entry into the system. Information in TEAPOTS provides some
insight as to the reasons why inquiries were not perfected during this
time period. In 20 percent of all inquiries for which TEAPOTS had data,
the complainant failed to respond; in 13 percent of the cases, the intake
staff found no valid basis for complaint; and in 8 percent of the cases,
the intake staff found no valid issue. However, TEAPOTS shows that 43
percent of the inquiries that were not perfected were coded as "Other
Disposition," which means that no further information is available to
indicate why the contact did not result in a perfected complaint.

In commenting on a draft of this report, HUD's General Deputy Assistant
Secretary for Fair Housing and Equal Opportunity stated that while HUD
considered requiring FHAP agencies to record initial inquiry dates for all
potential complaints, it has not done so because of the way HUD funds FHAP
agencies. Specifically, HUD reimburses the FHAP agencies for each (actual)
complaint that they investigate, and does not reimburse them for their
consideration of inquiries that do not result in complaints. However, this
continues to leave HUD without data or knowledge of a significant number
of potential Title VIII-related inquiries or a means of assessing the FHAP
agencies' response to such inquiries.

Some Logged Contact Dates Differed from TEAPOTS' `Date of First Inquiry'

Our prior report questioned the reliability of initial inquiry dates shown
in TEAPOTS. Our analysis for this report raises further questions, because
we found differences in the dates recorded in the contact logs and those
in TEAPOTS for the same contacts. HUD policy requires that FHEO offices
perfect or close all inquiries within 20 days of initial contact.
Statistics generated by HUD show an approximate 95 percent compliance with
this policy. However, for internal measurement and benchmarking purposes,
HUD begins counting the 20 days on the day the inquiry is entered into
TEAPOTS. Because HUD does not track the actual initial contact dates, it
can not use them to begin measuring the 20-day period. Our analysis of the
contacts recorded in the logs leading to the 306 perfected complaints
noted above indicates that 57 percent of the complaints were perfected
within 20 days of initial contact, based on the earliest contact date in
the log. However, using the inquiry date in TEAPOTS as the starting point
for the 20-day benchmark, as HUD does, indicates that 79 percent of the
complaints were perfected within 20 days. In fact, the median number of
days to perfect complaints was 11 using TEAPOTS inquiries, but 18 using
the data recorded in our log. These results indicate that HUD lacks an
accurate picture of how much time individuals face from the day they make
an inquiry to the day they learn the outcome of their cases, and that
HUD's reliance on TEAPOTS data leads to an inaccurate assessment of
performance in meeting its timeliness benchmark.

Case Files and TEAPOTS Lacked Evidence That Investigations Met Required
Standards or Followed Recommended Procedures

In reviewing investigative case files and associated TEAPOTS records, we
found that some lacked evidence that required investigative standards were
met, investigators followed recommended planning and procedure guidelines,
or that internal control measures were used and documented. Throughout
this section, we present estimates of agency compliance with certain
requirements and recommended practices based on our review of a random
sample of 197 FHAP agency and FHEO investigations that were closed during
the last half of 2004 either administratively, through conciliation, or
with a finding of no reasonable cause. Unless otherwise noted, these
estimates are surrounded by a confidence interval, due to sampling error,
of plus or minus 8 percentage points or smaller. We also present results
from our review of 12 of the 15 FHEO cases that completed the adjudication
process and subsequent monitoring during the same

period.29 Our review of each case was limited to reviewing the contents of
case files and the associated TEAPOTS records (that is, we did not
interview case investigators, other officials involved in the case,
complainants, or respondents), and it is important to note that the lack
of evidence we found does not necessarily indicate that required or
recommended steps were not taken. However, the lack of evidence does raise
questions about HUD's ability to assure that investigations are as
thorough as they need to be.

Some Files Lacked Evidence of Adherence to Investigative Standards

The Act sets several standards for investigators to follow during the
complaint process. First, investigators must establish and document four
jurisdictional elements to ensure that the complaint is covered under the
Act.30 Second, certain notifications to the complainants and respondents
must be sent and received.31 Third, a Final Investigative Report (FIR)
must be prepared at the end of each investigation.32 As a practical
matter, this means a FIR is required for investigations that conclude with
a determination of reasonable cause to believe a violation has occurred,
and for investigations closed with a determination of no reasonable cause
(i.e., there is no reasonable cause to believe a violation has occurred).
Investigators must also meet a 100-day deadline for completing an
investigation unless it is impracticable to do so.33 If the investigation
is not completed in 100 days, complainants and respondents should be
notified in writing of the reasons.34

Elements of Jurisdiction

Our review of investigative case files completed during the last half of
2004 showed that the elements of jurisdiction we measured were addressed
in nearly all of the no-cause cases  (see fig. 8).35 Similarly, we
observed that with one exception, all of the cause case files we reviewed
addressed the elements of jurisdiction we reviewed. The high incidence of
documentation addressing jurisdiction may be explained by the fact that
jurisdiction should be verified throughout the complaint process.
Consequently, there is more than one opportunity to identify deficiencies.
FHEO officials at one location said that in addition to an intake analyst,
the intake and enforcement or compliance branch chiefs also reviewed
complaints before investigations began. However, they also said that in
some instances, complaints for which jurisdiction had not been established
had been inadvertently accepted and later found to lack a required
jurisdictional element. For example, one case we reviewed showed that as
the investigation proceeded, investigators determined that the respondent
was exempt as a result of not owning more than one rental property. The
Act provides that certain properties and property owners are exempt, and
owners who do not have an interest in more than three single-family homes
or condominium units meet the guidelines for exemption.36

Figure 8: Percentage of Cases with Elements of Jurisdiction Addressed

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Notification

Although most files contained evidence that jurisdictional elements had
been addressed, we did not always find evidence that complainants and
respondents received required notifications. As noted above, the Act
requires the fair housing agency handling the complaint to send
complainants an initial notice acknowledging the filing of the complaint.
Further, HUD's complaint notification letter advises complainants of
certain guidelines of the complaint process.37 Respondents must be served
by certified mail or personal service with an initial notice of the
original complaint no later than 10 days after the complaint is filed or
when the respondent is identified. Respondents must also be notified
whenever a complaint is amended (complaints may be amended at any time
during an investigation to add or remove parties, and complainants must
sign the new complaints).38 Complainants and respondents should also be
notified when an investigation is closed. We therefore looked for evidence
indicating that these requirements were followed in all cases.

HUD regulations require that complainants be notified by certified mail or
personal service, but FHEO officials said that some FHAP agency procedures
do not require this. While we found initial notification letters addressed
to complainants in 97 percent of the files and letters to respondents in
91 percent for closure types other than reasonable cause, we frequently
did not find evidence that the letters had been received (see fig. 9). The
lack of evidence in case files that complainants and respondents had
received initial notifications does not necessarily mean that they did not
in fact receive the notices. Some FHEO officials told us that certified
mail receipts were sometimes maintained in separate files. We looked not
only for evidence such as return receipts from certified mail or
certificates of personal service, but also for correspondence indicating
receipt or knowledge of the complaint notification. Fifty-nine percent of
these cases contained evidence that complainants had received initial
notifications, and 67 percent contained evidence that respondents had
received initial notifications. Our survey of complainants that completed
the investigation process also indicates that some may not have received
initial notifications. Specifically, 86 percent said they received a
letter informing them whether an investigation would be conducted. We did
not observe a significant difference in documentation between
organizations. For the cases where HUD had determined reasonable cause, we
found initial notification letters addressed to complainants in 9 of the
12 files, and letters to respondents were found in 10. Evidence of receipt
was greater for reasonable cause cases than for other closure types-8 and
9 of the 12 files, respectively, had evidence that complainants and
respondents received initial notification letters.

Figure 9: Percentage of Case Files with Initial Complaint Notification
Letters Addressed to Complainants and Respondents and Evidence of Receipt

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Some cases did not contain evidence that final closure notices had been
addressed to complainants and respondents (see fig. 10). For closure types
other than reasonable cause, 10 percent and 21 percent of the files,
respectively, did not include copies of closure letters addressed to
complainants and all named respondents. Investigations conducted by FHAP
agencies were more likely not to have closure notices addressed to
respondents (26 percent) compared with FHEO-investigated cases (8
percent).39 Similarly, FHAP agency-investigated cases did not have
evidence of closure notices addressed to complainants 14 percent of the
time, compared with 2 percent for FHEO-investigated cases.40 Our survey of
complainants revealed that about 9 percent said they did not receive
notification of the case being closed. We found that for the cases where
HUD had determined reasonable cause, there were more notices of reasonable
cause determination addressed to respondents than complainants.
Specifically, 5 of the 12 files did not include notices addressed to
complainants informing them that the FHEO investigation was completed, and
of the reasonable cause finding. Two of the files did not include such
notices addressed to respondents.

For reasonable cause determinations, HUD's regulations require that all
parties to a complaint be notified of the reasonable cause determination
by certified mail or personal service.41 For no reasonable cause
determinations, the parties must be notified by mail and the notification
must include a written statement of the facts upon which the determination
was based.42 HUD guidance also states that for administrative closures,
all parties and their designated representatives must be notified by
regular or certified mail.

Figure 10: Percentage of Case Files with and without Closure Notices
Addressed to Complainants and Respondents (No Reasonable Cause)

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Fifteen percent of the case files for closure types other than reasonable
cause had evidence that the complaint had been amended. We also could not
find evidence in all cases that copies of the amended complaints had been
received by all respondents even though the statute requires that all

respondents receive them.43 Finally, not all of the cases, where
applicable, contained evidence that new respondents received a copy of the
complaint.44

As stated, FHEO officials noted that not all FHAP agency procedures
require certified mail notices, while HUD notifications are sent by
certified mail. As with initial notifications, we looked not only for
evidence of certified mail or personal service, but also other forms of
evidence that a notification was made, including response letters or
subsequent correspondence that indicated the parties' knowledge. FHEO
officials noted that the absence of notices in FHEO case files is more
likely a clerical omission than a failure to follow procedure.

Final Investigative Reports

Our file review showed that the Final Investigative Report (FIR) required
for reasonable cause and no reasonable cause outcomes and the
Determination showing the outcome of cases were not always present. The
documents are intended to demonstrate that investigations were thorough
and that the investigator's conclusions were founded in fact and evidence.
FIRs, which HUD guidance states fulfill the statutory requirement to
document investigations, are used as a basis for preparing the charge for
reasonable cause cases. FIRs should summarize the allegation and evidence,
including such things as dates and summaries of contacts with parties,
witness statements, descriptions of pertinent records, and answers to
interrogatories.45 The Act requires FHEO, following the completion of the
investigation, to make information derived from the
investigation-including the FIR-available upon request to the parties to
the complaint.46 The Determination includes the elements of jurisdiction
as well as a summary of the complainant's allegations, the respondents'
defenses, and the investigator's findings and conclusions.

For cases where HUD had determined reasonable cause, we found the FIR and
Determination in all 12 of the files. However, for cases closed with a
Determination of no reasonable cause, we found the FIR missing in 5
percent of the files, and a Determination missing from 8 percent.47 The
percentage of FIRs for no reasonable cause cases was similar for FHAP
agency and FHEO files.

HUD requires that FIRs and Determinations of reasonable cause and no
reasonable cause be approved by the FHEO regional director. FHAP agency
managers approve and sign these documents for FHAP agency investigated
cases. For cases where HUD had determined reasonable cause, we found that
11 FIRs and 10 Determinations had been signed. For cases with no
reasonable cause outcomes, 71 percent of Determinations were signed,
compared with only 45 percent of FIRs.48 For no reasonable cause outcomes,
FHEO files showed more evidence of Determination approval-100 percent
compared with 60 percent of FHAP agency Determinations.49 FHEO noted that
missing signatures for FIRs and Determinations are more likely an
oversight rather than a question of thoroughness or lack of review. They
also noted that case files will not document informal means of review.

Adherence to 100-Day Notice Requirement

As noted above, the Act requires that fair housing investigations be
completed within 100 days from the date the complaint was filed, unless it
is impracticable to do so. An investigation is completed when a
Determination or charge is issued, a conciliation agreement is executed,
or the complaint is otherwise closed.  50 We estimate that 98 percent of
all cases with closure types other than reasonable cause-including
relatively more no reasonable cause cases-took more than 100 days to
complete. If investigators do not meet the 100-day deadline, the
investigating agency is required to notify both complainants and
respondents in writing,

explaining why the investigation is not complete.51 We found 100-day
letters in each file for these closure types with investigations that took
more than 100 days to complete. We found that in about two-thirds of these
cases, the 100-day letter was sent after 100 days had passed.52 Moreover,
about 14 percent of the notices we found were dated more than 130 days
after the HUD filing date (see fig. 11).53 For cases where HUD had
determined reasonable cause, all 12 investigations lasted longer than 100
days, but 2 of the files did not have copies of 100-day letters.

Figure 11: Timing of 100-Day Notices Sent to Complainants for
Investigations Lasting over 100 Days

Note: Due to rounding, total may not equal 100 percent.

According to FHEO officials in one location, for cases with complex
issues, it was often difficult to meet the 100-day investigative
requirement and conduct a thorough investigation. Officials in another
FHEO location said that the 100-day time frame is a critical factor at day
one, and a new initiative had been implemented to track and focus on cases
at the 50-day mark. The FHAP agencies' record of meeting the 100-day
requirement is directly tied to their performance ratings and to the
reimbursement they receive for completed cases. Officials at one FHAP
agency stated that the 100-day requirement was a priority for each new
investigation and their agency had established shorter investigative
deadlines internally, using their own streamlined process to assist them
in meeting the 100-day requirement. In our last report, we observed that
the percentage of investigations completed within 100 days increased
between 2001 and 2003, particularly for FHEO cases. Specifically, the
percentage of FHEO investigations completed within 100 days increased from
17 percent in fiscal year 2001 to 50 percent in fiscal year 2003. We also
noted that FHEO hub directors reported that the 100-day benchmark and the
simultaneous need to conduct a thorough investigation were sometimes
competing goals.

Some Files Lacked Evidence That Investigators Followed Recommended
Procedures

In addition to the statutory requirements, HUD guidance recommends a
number of activities that contribute to a more complete investigation.
Among these are preparing investigative plans, conducting on-site visits
and interviews, and requesting policy and procedure information from
respondents.54 The guidance also recommends that investigators follow
certain procedures before closing a case administratively. HUD officials
said that an investigation may be thorough without following each
recommended practice; moreover, as noted above, a lack of evidence does
not necessarily indicate that a procedure was not followed. However, the
relative infrequency with which some practices were used, according to
documentation in the case files, raises questions about investigation
thoroughness.

Investigative Plans

HUD guidance states that investigative plans are critical to efficient and
effective investigation. The guidance also provides extensive instruction
on preparing plans, and adds that, in developing investigative plans,
investigators and their supervisors should consult with Regional Counsel.
According to HUD's revised Title VIII Handbook, investigative planning
allows supervisors and investigators to ensure that the scope of the
investigation is carefully tailored for adequate investigation of all
claims made in the complaint, and careful planning should also prevent
"over-investigation" of claims. However, FHEO officials stated that most
experienced investigators do not prepare investigative plans for any
except technical and very complex cases since investigators are familiar
with the procedures for more common discrimination cases.

Our file reviews showed that 62 percent of the cases with closure types
other than reasonable cause did not include investigative plans. Further,
the plans we found were not as detailed as the guidance suggests. For
example, while the type of discrimination was identified in 93 percent of
plans, the theory of discrimination was identified in 65 percent.56 Also,
55, planned interviews with complainants and respondents were specified in
59 percent and 62 percent of plans, respectively, and a list and sequence
of interview questions was present in 27 percent of plans (see fig. 12).57
For cases where HUD had determined reasonable cause, we found an
investigative plan in 10 of the 12 files, but these also contained little
detail. For example, the type of discrimination was identified in all 10
of the plans, but the theory of discrimination was identified in only 3.
Additionally, only 2 and 3 of the plans, respectively, specified
interviews with complainants and respondents, and a list and sequence of
interview questions was present in only 2 of the plans. With regard to
supervisory review, officials at the agencies we visited stated that there
was no established procedure for documenting review of investigative
plans.

Figure 12: Percentage of Cases with Investigative Plans and Cases with
Specific Content in Investigative Plans

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

We found substantially more investigative plans in FHEO case files (74
percent) than in FHAP files (24 percent). There was little variation in
the percentage of cases with investigative plans across closure types (see
fig. 13).

Figure 13: Percentage of Cases with and without Investigative Plans, by
Investigating Agency and Closure Type

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Interviews with Complainants and Respondents

HUD's training manual for fair housing investigators states that
interviews are a vital part of collecting evidence, and that they allow an
investigator to probe for additional information that otherwise might not
be provided. The manual recommends that investigators interview not only
complainants and respondents, but also other individuals, such as
witnesses for the parties and independent witnesses. FHEO officials noted
that there are circumstances where complainant or respondent interviews
are not necessary, such as when a case is conciliated before the
investigation has begun or when an investigator determines that a
respondent is exempt under the Act.

The majority of cases with closure types other than reasonable cause
included interviews with the parties to the complaint, but 28 percent did
not include interviews with respondents. Cases with no reasonable cause
outcomes were more likely to include interviews with complainants and
respondents (see fig. 14). FHEO cases for these closure types were more
likely than FHAP cases to include interviews with complainants.
Specifically, we estimate that FHEO cases did not include interviews with
complainants 8 percent of the time, compared with 21 percent of the time

for FHAP cases.58 We found at least one respondent interview for all cases
where HUD had determined reasonable cause. By contrast, only seven of
these cases included interviews with complainants.

Figure 14: Percentage of Cases with and without Complainant and Respondent
Interviews, by Closure Type

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

In a number of cases, investigators interviewed complainants and
respondents once or twice. Forty-seven percent of cases with closure types
other than reasonable cause showed one or two complainant interviews, and
40 percent showed one or two interviews with a respondent or
representative. We found that the content of interviews recorded in the
TEAPOTS database varied widely.

We looked specifically for evidence explaining why interviews were not
conducted or whether it was apparent why interviews were not conducted.
While we often could not find documentation as to why complainants and
respondents were not interviewed, the reason was apparent in a number of
cases.

We also looked at the time frame within which the parties were interviewed
for the first time following the official HUD filing date. The 100 days
that HUD allocates for conducting fair housing investigations begins on
the date a complaint is officially filed with HUD. HUD's training manual
for fair housing investigators suggests that investigators interview
complainants first to clarify the details of the allegation and to
evaluate the viability of the complaint. We found that complainants were
typically interviewed first, but that in a number of cases, initial
investigative interviews were conducted weeks and even months after the
complaint had been filed.59 For closure types other than reasonable cause,
65 percent of cases with complainant interviews showed the first
investigative interview occurring more than 2 weeks after the complaint
was filed; this was also the case for 67 percent of respondent
interviews.60 Further, 52 percent of cases with complainant interviews
showed the first investigative interview occurring more than 1 month after
the complaint was filed; this was also the case for 46 percent of
respondent interviews.61

FHEO officials noted that it may be appropriate to conduct a respondent's
initial interview after first receiving documentation. Respondents have 10
days to provide a response to the complaint. FHEO officials noted that in
addition to caseload, the time involved in mail delivery and the
difficulty of locating people who have moved are factors that can impact
the initial interview with the parties. Officials at some of the FHAP
agencies we visited indicated that time may be of the essence in housing
discrimination cases since housing opportunities may be lost during the
initial filing period. For pending evictions, one agency's practice was to
request an eviction abeyance through the court or to petition the landlord
to postpone eviction pending a speedy investigation. To expedite the
filing process, the intake coordinator at another agency may go on site to
have the complaint form completed and signed in order to begin the
investigation.

On-Site Visits

HUD's guidance does not require that an investigator visit the site of the
alleged discriminatory act, such as the subject dwelling and respondent's
place of business, and FHEO officials stated that many cases do not
require an on-site visit. However, the guidance also states that an
on-site visit is the most efficient way to conduct an investigation in
most situations involving fair housing complaints. Additionally, it states
that while such a visit may not appear necessary at the beginning of an
investigation, issues may develop as a case progresses that support the
necessity of an on-site investigation. One FHAP agency we visited had a
policy of conducting on-site interviews or inspections for each
investigation, and officials there reported that 95 percent of their cases
included visits to the property in question to conduct interviews and to
obtain supporting documentation. In contrast, officials at FHEO offices
and other agencies said that limited financial resources and time
constraints may prevent investigators from including on-site visits as a
routine part of their investigations.

About three-quarters of the cases we reviewed with closure types other
than reasonable cause showed no evidence that investigators made on-site
visits. For non-cause cases that included an on-site visit, we found that
investigators toured the property in question, collected physical evidence
such as photographs, and interviewed complainants and respondents (see
fig. 15). We saw no statistically significant difference in use of on-site
visits among the non-cause closure types, but we found substantially more
on-site visits documented for cases where HUD had determined reasonable
cause. Ten of these cases included on-site visits, and investigators
generally carried out more activities while on site. The investigator
toured the property in 4 cases, physical evidence was collected in 9, and
both complainants and respondents were interviewed in 7 of the 10 cases.

Figure 15: Percentage of Cases with On-Site Visits and Investigative
Activities Conducted during On-Site Visits

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Information Requests

HUD's training manual for fair housing investigators recommends that
investigators request information on respondents' policies and procedures
to compare the established policies and procedures with the alleged
discriminatory practice. Policy and procedure documents may take a variety
of forms, including lease agreements and housing covenants. We found that
74 percent of cases for closure types other than reasonable cause
contained evidence that investigators requested policy and procedure
information from respondents. Although policy and procedure documents are
not always necessary to establish reasonable cause, we found that such
documents were requested in all of the cases where HUD had determined
reasonable cause.

The manual also suggests that investigators request comparative
information, especially in cases alleging unequal treatment, about persons
in the same protected class as the complainant and persons not in the

complainant's protected class.62 We found that the files for 58 percent of
cases for closure types other than reasonable cause showed that
comparative information was requested. FHEO officials noted that for cases
involving refusal to rent, they generally would expect that investigators
collect such information. However, in cases involving design and
construction for access by persons with disabilities, such information
typically is not necessary. For cases with closure types other than
reasonable cause, we found that 30 percent of cases where refusal to rent
was at least one issue did not show evidence that comparative information
was requested. For cases where HUD had determined reasonable cause, we saw
evidence in 10 of the 12 files that comparative information was requested.

Comparative information was requested most often for cases with a finding
of no reasonable cause (see fig. 16).63 Specifically, we estimate that 82
percent of no reasonable cause cases had comparative information
requests.64 FHEO officials believe that recommended practices for planning
investigations, interviewing complainants and respondents, conducting
on-site visits, and seeking policy information need not be carried out in
every case. They believe that every case is unique and each investigation
should be tailored to the case. Our prior report noted that at least one
FHAP agency had developed software that automatically generated a list of
critical documents that were usually needed for certain types of
investigations. According to officials of this FHAP agency, this system
improved the quality of investigations and decreased the length of cases.

Figure 16: Percentage of Cases with Policy and Procedure Information
Requests, by Closure Type

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Attempts to Reach Complainants Prior to Administrative Closure

Finally, we found that investigators documented multiple attempts to reach
complainants before closing complaints administratively. Twenty-one
percent of the cases we reviewed had been closed administratively, most of
them because complainants withdrew their complaints. Relatively few of the
cases closed administratively resulted because a complainant could not be
located or was uncooperative with the investigator. Nonetheless,
investigators documented as many as 11 attempts to contact uncooperative
complainants by telephone, certified mail, and regular mail. The
investigating agency is required to notify the parties of administrative
closures as with other closure types, and 82 percent of administrative
closures had evidence that such notices had been addressed to the
parties.65

TEAPOTS Was Used Inconsistently to Record Investigation Activity

Selected information entered in TEAPOTS was generally consistent with the
information found in source documents in the case files, but use of the
system varied considerably among agencies conducting fair housing
investigations. Complete and reliable TEAPOTS information is important for
each case since the database is used to record activities and information
throughout the investigation and subsequently as a resource for preparing
investigative reports. In addition, HUD officials point to TEAPOTS as a
control to determine that investigations are conducted in accordance with
statute and regulation. Without including evidence as an investigation is
completed, TEAPOTS cannot provide an accurate representation of the
evidence.

As part of our review of case files, we traced information in TEAPOTS
relating to the basis and issue of discrimination, the date of the last
alleged violation, and the HUD filing date and found that it generally
matched source document information, with some exceptions. We found that
for more common complaint issues-discriminatory terms and refusal to
rent-the information matched 83 percent and 81 percent of the time,
respectively.66 Similarly with regard to the discrimination basis, the
information matched 92 percent and 84 percent for race and national
origin, respectively, but the evidence is less clear for color.67 For
familial status, the information matched 86 percent of the time.68 When we
compared the date of the last alleged violation and the HUD filing date in
TEAPOTS with the same dates in the source documents, we found that these
matched 91 percent and 84 percent of the time, respectively.69 Also, an
estimated 91 percent of the last alleged violation dates in TEAPOTS
matched the source documents. Finally, filing dates in TEAPOTS matched
those in the source documents in 84 percent of cases.

In reviewing TEAPOTS reports for investigation details, we saw that the
amount of information varied a great deal depending on which agency had
investigated the case and entered the information. In some instances,
entire sections such as those for recording interviews, case chronology,
and the investigator's findings and conclusion had not been completed.
Without a complete record of an investigation, investigators may be unable
to utilize TEAPOTS functions for preparing investigative reports. Further,
because the statute requires that information derived from investigations,
including the FIR, be made available to complainants and respondents upon
request, it is pertinent that all appropriate details be included.

Evidence Indicates a Lack of Consistent Efforts to Conciliate Complaints

Although HUD requires fair housing agencies to attempt conciliation
throughout the complaint process, our review of case files, survey of
complainants, and test calls revealed that FHEO and FHAP agencies did not
always attempt to conciliate complaints, made limited efforts to do so, or
did not meet HUD's requirement that they document these efforts. While
having the fair housing specialists act as both investigators and
conciliators is permitted, investigators faced with pursuing conciliation
as well may focus on investigative activities, particularly considering
FHEO's emphasis on completing investigations within 100 days.

Conciliation Was Not Always Discussed during Intake

FHEO's Title VIII Handbook states that conciliation should be discussed
during the initial intake interview and should be noted in the standard
notification letters to the complainant and respondent. Further, the
statute requires that conciliation must be attempted to the extent
feasible commencing with the filing of the complaint and concluding with
the issuance of a charge on behalf of the complainant, or upon dismissal
of the complaint. For time-sensitive complaints, conciliation may be the
most effective procedure and, given that the resources of FHEO and FHAP
agencies are limited, an effective means of reducing staff workloads.
FHEO's General Deputy Assistant Secretary for Fair Housing and Equal
Opportunity noted that conciliation is an integral part of the complaint
process. He noted that FHEO officials have "confidence that every party is
informed during the initial interview or contact of HUD's statutory
obligation to attempt conciliation and is asked about the possibility of
conciliating the complaint and what it might take to effect conciliation."

According to our survey of complainants, fewer than half (42 percent) were
offered assistance with conciliation (see fig. 17).70 In 26 percent of the
cases, complainants said staff suggested that the parties work out their
differences on their own.

Figure 17: Complainant Views on Fair Housing Complaint Staff Conciliation
Assistance

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Further, during the test calls we placed to FHEO and FHAP agencies, we
found the possibility of conciliation was discussed in only 18 percent of
the locations we called. FHAP agencies mentioned conciliation 23 percent
of the time, and FHEO did not mention conciliation at all.

Based on our survey of complainants, we estimate that nearly 90 percent of
complainants who were offered conciliation accepted it. About 12 percent
of complainants said they sought help with conciliation through other
organizations. The percentages did not vary among cases depending on
whether they were investigated by FHEO or a FHAP agency, but did vary
based on closure type, with approximately half of the complainants who
were offered conciliation far more likely to have their case end with a
conciliation outcome. We estimate that complainants with conciliation
outcomes were offered conciliation 67 percent of the time, while those
with no-cause outcomes were offered conciliation just 27 percent of the
time.

Overall, we estimate that parties involved in fair housing complaints were
more likely to reach conciliation agreements when FHEO or a FHAP agency
was involved. For example, we estimate that complainants reached agreement
64 percent of the time when FHEO or a FHAP agency assisted with
conciliation, but only about 35 percent of the time when another
organization did the conciliation. Complainants who worked with FHEO or a
FHAP agency were satisfied with the outcome about 81 percent of the time
and with the other party's compliance with the agreement more than 90
percent of the time, compared with 47 percent and 90 percent,
respectively, for complainants working with other organizations.

Our previous report noted that one FHAP agency was experimenting with a
separate "mediation track" when handling complaints. At this agency,
mediation attempts occurred early in the process during intake and
involved a professional private-sector mediator. The mediation had usually
pleased the parties, resulting in timely resolutions of cases and
beneficial results. Two FHAP agencies we visited for this study offered
mediation. Further, other FHAP agencies had begun offering mediation
during the intake stage, and the complainants' decision to participate or
not participate in mediation was almost always documented. One FHAP in its
notification letter to the complainant offered the complainant the choice
of two options-either conciliation or investigation. Staff from one FHEO
regional office noted the use of mediation by a FHAP agency in their
region reduced complaints by approximately one-third.

Case Files Often Contained No Documentation of Conciliation Attempts

During our file review of no-cause cases, we found no documentation of
conciliation attempts around a third of the time and often no
documentation of contacts with either party to attempt conciliation.
Specifically, an estimated 36 percent of the files contained no evidence
that complainants had been contacted to attempt conciliation, while 30
percent of files lacked evidence that respondents had been contacted to
attempt conciliation. For the 12 cases with outcomes of reasonable cause
that we reviewed, we found documentation that conciliation was attempted
in 11 of them.

When an agency contacted parties to a complaint, they often did so once or
twice. For example, we estimate that 21 percent of complainants in cases
with no-cause outcomes were contacted only once.71 As indicated in figure
18, we also found that conciliation attempts varied somewhat by closure
type. Conciliation was attempted with complainants less often for cases
that were closed administratively. However, for all cases, we found that
information on conciliation contained in the case file varied
tremendously, with some cases noting "conciliation discussed," while
others included significant details.

Figure 18: Documented Conciliation Attempt, by Closure Type

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

As is the case with other aspects of our file review, the lack of evidence
of conciliation attempts does not necessarily indicate that such attempts
did not occur; attempts may not have been documented. However, according
to FHEO's Title VIII Handbook, the lack of detailed documentation of
conciliation attempts could be problematic. For example, the Handbook
noted that FHEO Headquarters or the Office of General Counsel occasionally
questioned the sufficiency of conciliation efforts for cases forwarded to
FHEO Headquarters with recommendations of a Determination of reasonable
cause. Further, it stated that when these questions were asked, the case
was often remanded to the agency handling the case with instructions to
undertake additional late-stage conciliation efforts. The Title VIII
Handbook noted that a number of these remands result not from a lack of
initial conciliation attempts, but rather from a lack of documentation of
conciliation attempts in the case file.

Conciliation attempts should be documented in TEAPOTS before a case can be
closed. TEAPOTS includes a separate section that allows the intake
specialist and investigator to document conciliation efforts. While HUD
relies on TEAPOTS as a control to assure that investigations, including
conciliation attempts, are performed in accordance with statute and
regulation, information that we obtained from TEAPOTS shows that its use
varied from one location to another. For example, the descriptions of
conciliation attempts varied in detail, and the case information recoded
by some FHEO and FHAP offices did not include a chronological listing of
conciliation attempts as suggested by the Title VIII Handbook.

Finally, we found that conciliation agreements were generally
well-documented. Specifically, an estimated 91 percent of cases closed
with conciliation included copies of conciliation agreements in the case
file.72 A HUD conciliation agreement is a written, binding agreement to
resolve the disputed issues in a Title VIII housing discrimination
complaint. The HUD conciliation agreement must contain provisions to
protect the public interest in furthering fair housing.73 According to the
Act, a conciliation agreement requires HUD approval.74 Approximately
ninety-five percent of these agreements were signed by all parties and
approximately 90 percent were approved by FHEO or the FHAP agency.75

Complainants Viewed Conciliation as a Case Outcome More Favorably Than
Other Possible Outcomes

Based on our survey, complainants whose cases were conciliated were more
positive with this outcome than those who experienced administrative
closure or a finding of no-cause, but complainants felt at least some
pressure to conciliate their complaints, most commonly because they felt
their cases would not be handled otherwise. As shown in figure 19, 52
percent of the complainants surveyed indicated that they felt a little to
great pressure to resolve their cases. In commenting on a draft of this
report, the General Deputy Assistant Secretary for Fair Housing and Equal
Opportunity noted that the Fair Housing Act mandates attempts at
conciliation and that this statutory construct may result in what
complainants perceive to be pressure to resolve their case.

Figure 19: Most Commonly Cited Pressures to Resolve Case

Note: Brackets on each bar represent the sampling error (confidence
interval) for that estimate.

Our survey of complainants indicated that conciliation seemed to work when
an agency such as FHEO or FHAP agencies were involved (over 64 percent
resulted in a conciliation). However, a significant number of survey
complainants indicated that this opportunity was not presented to them.
Moreover, those who sought conciliation assistance from any other
organization were less likely to reach a satisfactory outcome; 47 percent
did not realize a satisfactory agreement.

Emphasis Given to Conciliation May Be Affected in Cases of Dual
Investigator/Conciliator Role

Our previous report noted that investigators at some FHEO locations and
FHAP agencies customarily conciliated their own cases, while other
locations usually used separate investigators and conciliators.76 Also,
our previous report noted that officials were divided on the impact of
this practice. Some officials told us that having the same person
performing both tasks had not caused problems. Other officials-including
some at locations where investigators conciliated their own
cases-indicated a preference to have different people perform these tasks.
One official said that separating these tasks enabled simultaneous
conciliation and investigation of a complaint, a practice that speeded up
the process. Another official noted that parties might share information
with a conciliator that they would not share with an investigator and that
a conflict of interest might result if one person tried to do both. The
same official said that although investigators were not allowed to use
information they learned as conciliators during investigations, the
information could still influence the questions conciliators posed-and
thus the information they learned-as investigators. Similarly, at one FHEO
location hub, an OGC official told us that information learned as a result
of conciliation efforts should not be included in investigative findings.
A few enforcement officials at locations that did not separate the
functions said they did not have enough staff to have separate
conciliators. We recommended in our previous report that FHEO establish a
way to identify and share information on effective practices among its
regional fair housing offices and FHAP agencies.

According to the Title VIII Handbook, conciliation may be a fertile source
of information regarding a respondent's housing practices. However, the
Handbook notes that nothing said or done during the course of conciliation
can be used in the investigator's reasonable cause recommendation, in the
final investigative report, or in any subsequent Title VIII enforcement
proceeding. Information discovered during conciliation should not be made
public without the written consent of the persons concerned. Although
information discovered during the conciliation process cannot be factored
into the investigator's recommendation, if this same information is
discovered outside of the conciliation process, it is permissible for
investigators to use this information in their recommendations. For
example, if respondents make an admission during conciliation
negotiations, investigators cannot use this admission in their
recommendations. However, if respondents make this same admission in a
later deposition, the investigator can use this admission in their
recommendations.

In our previous report, we also noted that some HUD locations we visited
put investigations on hold when conciliation looked likely, while others
did not.77 Some fair housing officials at the locations that
simultaneously investigated and conciliated cases told us that doing so
not only expedited the enforcement process but could also facilitate
conciliation. Because the parties were aware that the investigation was
ongoing, two FHEO hub directors told us, they were sometimes more willing
to conciliate. Additionally, some officials at the offices that delayed
the investigation while attempting conciliation told us that this practice
increased the number of calendar days necessary to investigate a case.
However, one FHEO hub official told us that simultaneous investigation and
conciliation could waste resources, as it might not be necessary to obtain
further evidence in a case that would be conciliated. Overall, 6 of the 10
hub directors told us that simultaneous investigation and conciliation had
a great or very great impact on the length of the complaint process, and
all 6 said that the practice decreased the length.

During our current review, officials from one FHEO regional office noted
that using separate conciliators would definitely assist in making their
investigative process more effective. However due to staffing constraints,
they believed it was impractical to do so without a significant increase
in staff. These officials noted that they needed additional staff to speed
up the investigative process, separate investigation from conciliation,
conduct more thorough investigations, and more effectively monitor
compliance with conciliation agreements.

Complainants Were Dissatisfied with the Fair Housing Complaint Process,
Its Outcome, and Certain Aspects of Intake and Investigation

While an estimated 44 percent of complainants were somewhat or very
satisfied with the fair housing complaint process, based on our survey, we
estimate that about half of all complainants were either somewhat or very
dissatisfied.78 Similarly, nearly 60 percent were dissatisfied with the
outcome of the fair housing complaint process, and almost 40 percent would
be unlikely to file a complaint in the future. Complainants'
dissatisfaction varied for each stage of the complaint process, as well as
by type of complaint closure (administrative, conciliation, and no-cause
finding), with complainants in no-cause cases expressing the least
satisfaction with various aspects of investigations. However, according to
FHEO, the low satisfaction levels of complainants with a no-cause finding
is not wholly unexpected given that they failed to receive the desired
outcome and thus question the process that produced the outcome.

Many Complainants Were Dissatisfied with the Complaint Process, but
Satisfaction Varied According to the Stage of the Process and Case Outcome

Overall, about 34 percent of all complainants were satisfied with both the
process and the outcome; conversely, 48 percent of all complainants were
dissatisfied with both the process and the complaint outcome (see fig.
20). When looking at complainants' overall satisfaction level, we found no
significant differences when we sorted by type of agency--that is, between
cases investigated by FHEO and those investigated by FHAP agencies.
However, variations by closure type existed. For example, over two-thirds
of those with a no-cause finding were dissatisfied with both the process
and the outcome of their complaint (about 68 percent). In contrast, over
two-thirds of those closed through conciliation were satisfied with both
the process for handling their complaint and its outcome (about 68
percent). Also, just under half of the complainants with administrative
closures were dissatisfied with both the process and outcome (about 43
percent).

Figure 20: Proportion of Complainants Who Were or Were Not Satisfied with
the Overall Complaint Process and the Outcome (Percentage of all
Complainants)

Note: Totals do not equal 100 percent because figure does not include
"don't know" responses.

While an estimated one-half of all complainants-regardless of their case
outcomes--were either somewhat or very dissatisfied with their experience
with the overall complaint process, the percentages expressing
dissatisfaction with the intake and investigative stages were smaller (see
fig. 21). Specifically, about 71 percent of complainants were somewhat to
very satisfied with the intake stage and nearly 55 percent with the
investigative stage.

Figure 21: Complainant Satisfaction with the Overall Process and the
Intake and Investigative Stages

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

In addition, we found that complainants who were dissatisfied with the
outcome of the complaint process were also likely to express
dissatisfaction with the process itself. These results were consistent
across FHEO offices and FHAP agencies. Further, complainants with a
no-cause finding were more likely to be dissatisfied with both the process
and outcome than complainants whose complaints were conciliated or closed
administratively. For example, complainants with a no-cause outcome were
somewhat to very dissatisfied with the process 72 percent of the time (23
percent were satisfied) and with the outcome of their cases 84 percent of
the time, while those with a conciliation outcome reported dissatisfaction
levels of 21 percent with the process (75 percent were satisfied) and 25
percent with the outcome.

About 40 percent of complainants said they would be somewhat to very
unlikely to file any future complaint with the same fair housing agency.
These results did not differ significantly by type of agency but did
differ by closure type. For example, complainants with a no-cause outcome
said they would be somewhat to very unlikely to file another complaint
about 56 percent of the time, compared with 14 percent of those whose
cases were conciliated.

Generally, complainants' level of satisfaction with the process and its
outcome did not vary with the expectations they had before talking to
anyone at a fair housing organization. There were a few important
exceptions, however. About 20 percent of complainants reported that they
expected that the fair housing organization would not help both sides
equally. These complainants were significantly more dissatisfied with the
overall process and its outcome. The same was true for the approximately
30 percent of complainants that expected the fair housing organization
would get the complainant a financial award. Half of all complainants had
expectations that were not listed in our survey.79 Those complainants that
had expectations other than those we listed were significantly more
dissatisfied with each stage of the process, the overall process, and its
outcome. These results may indicate that some complainants have different
or greater expectations than others.

Complainants Viewed the Overall Intake Process Positively but Expressed
Concern about Specific Intake Activities

According to our survey, about 71 percent of complainants were somewhat or
very satisfied with the intake process (see fig. 22). More than half of
the complainants reported that they received clear information during the
intake process and that intake staff were courteous and mostly acted
promptly. Yet, a substantial number of complainants gave poor ratings to
specific aspects of the process, citing difficulty contacting intake staff
and the lack of timeliness of some intake activities. These opinions were
generally true across agency type and closure types.

Figure 22: Complainant Satisfaction with Intake Stage, by Closure Type

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

Complainants reported that intake staff at both FHEO and FHAP agencies
provided understandable information more than half of the time, including
satisfactory explanations about an agency's decision on whether to pursue
an investigation. In general, about 60 percent of the time complainants
told us that they received clear information on the likely length of each
step in the process and explanations of the complaint and investigative
process. Moreover, an estimated 66 percent of complainants were very or
somewhat satisfied with the way the organizations explained their decision
to pursue or not pursue a case (see fig. 23). The results varied by
closure type with the exception of the organization's explanation of their
decision whether to investigate. For example, fewer complainants with
no-cause outcomes, relative to complainants with conciliation outcomes,
felt that they had received understandable information on the time
involved or felt that they had received explanations of the complaint and
investigative processes.

Figure 23: Complainant Views on Whether Intake Staff Provided
Understandable Information and Satisfactory Explanation on Whether the
Complaint Would Be Investigated

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

Complainants Reported That Intake Staff Took Prompt Action Most of the
Time and Were Generally Professional and Courteous

Based on survey results, complainants believe intake staff took action in
a timely manner more than half the time on the intake activities we
reviewed. Specifically, we estimate that about 70 percent of the time
intake staff sought the complainant's signature somewhat to very quickly
after the initial contact and that 62 percent of the time intake staff
acted somewhat to very quickly in deciding whether to pursue an
investigation (see fig. 24). Staff performance in getting back to
complainants was apparently less satisfying, with just 55 percent of
complainants responding that intake staff acted somewhat to very quickly,
a result that complements our findings on the difficulty of contacting
intake staff. In general, these results were true regardless of whether
FHEO or FHAP agencies had handled the complainants' cases. Complainants
with no-cause outcomes cited the slowest response time, saying that staff
responded somewhat or very quickly only about 40 percent of the time. On
the other hand, complainants whose cases were conciliated reported the
fastest agency action for certain actions.

Figure 24: Complainant Views on Timeliness for Certain Intake Activities,
by Closure Type

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

In general, complainants felt that intake staff provided services in an
acceptable and professional way (see fig. 25). We estimate that intake
staff at both FHEO and FHAP offices treated complainants with courtesy and
respect about 85 percent of the time and were helpful and impartial more
than 70 percent of the time, according to complainants we surveyed.
Complainants with no-cause outcomes reported positive treatment less
often. For example, these complainants said that intake staff were helpful
and interested in their complaint about 60 percent of the time and
thorough about half of the time.

Figure 25: Complainant Views on Treatment by Intake Staff

Note: Totals do not equal 100 percent because figure does not include
"don't know" responses. Brackets on each bar represent the sampling error
(confidence interval) for that estimate.

Complainants Reported That Intake Staff Performed Some Activities More
Often and More Quickly Than Others

We asked complainants whether intake staff carried out a variety of
activities that were either required or that could be considered best
practices-for example, did intake staff notify complainants whether the
fair housing organization would pursue the case? According to survey
results, FHEO and FHAP agency staff carried out some intake-related
activities, such as seeking signatures for a complaint, more frequently
and more quickly than others, such as taking action to prevent the loss of
a housing opportunity. Among other things, our survey showed the
following:

o Intake staff very often asked complainants to sign a complaint. This was
true for about 90 percent of complainants who were working with a single
fair housing agency. In contrast, only about 48 percent of those
complainants who were working with two or more fair housing agencies were
asked to sign a complaint.80 We did not observe any significant
differences by type of agency (FHEO or FHAP) or closure type. Because we
surveyed only complainants that had filed complaints, we would expect that
all would have been asked to sign the complaint. According to the Title
VIII Handbook, a complaint should generally be signed before it can be
considered as filed, so as to provide protection against frivolous or
false claims or inadvertent erroneous statements on the intake form.

o Complainants stated by a large margin-about 81 percent-that intake staff
asked questions that would help the agency understand what led to the
complaint, and 86 percent stated that the agency notified them with a
decision on whether an investigation would be undertaken. Again, all
complainants should be asked questions about their allegation, including
information needed to satisfy the required elements of jurisdiction. In
general, these results were similar across FHEO, the FHAP agencies, and
the closure outcomes.

o Complainants reported that intake staff were less likely to take some
actions or to ask certain questions (see fig. 26). For example, according
to complainants we surveyed, about 69 percent of the time both FHEO and
FHAP intake staff did not attempt to prevent the loss of a housing
opportunity when asked to do so, although the percentage varied across
outcome types. Complainants with administrative closures or conciliated
cases were slightly more likely to report that staff took action for them
than those with a no-cause outcome.

o Complainants also said that intake staff did not offer to resolve
differences between parties about 45 percent of the time. Again, the
results differed according to case outcome, with complainants who
conciliated reporting the most offers (76 percent) and those with a
no-cause outcome the least (37 percent). As previously discussed, the Act
requires the fair housing organization to offer conciliation to the extent
feasible in all cases.

Figure 26: Complainant Views on Whether Intake Staff Carried Out Certain
Activities

Note: Totals do not equal 100 percent because figure does not include
"don't know" responses. Brackets on each bar represent the sampling error
(confidence interval) for that estimate.

Most Complainants Reported Being Satisfied with the Investigative Stage,
but Substantial Numbers Rated Poorly Certain Aspects of the Investigation
Process

Although most complainants were satisfied with the investigative stage of
the complaint process, they were generally less positive than they were
about the intake stage. Further, a substantial number of complainants
expressed dissatisfaction and concern about certain aspects of
investigations.

We estimate that about 40 percent of complainants were dissatisfied with
the conduct of investigations, whether it was conducted by FHEO or a FHAP
agency (see fig. 27). As they were with other activities, complainants
whose cases were closed with a no-cause outcome were the most dissatisfied
with the conduct of investigations, with nearly two-thirds reporting
dissatisfaction. The concerns that led to this dissatisfaction included
problems contacting staff, concern that staff did not perform actions such
as informing complainants about options after their case was closed, and
difficulty obtaining clear information. However, those whose cases were
conciliated very often reported being satisfied with the investigation.
Despite the concerns cited, complainants in general believed that staff
treated them professionally and with respect and courtesy.

Figure 27: Complainant Satisfaction with Investigative Stage, Overall and
by Closure Type

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

We estimate that a quarter to a third of complainants had problems
reaching investigators, believed that investigators performed poorly in
providing case updates, and were dissatisfied with the amount of contact
they had with investigators. An estimated one-quarter of the complainants
found it hard or somewhat hard to reach investigators, and more than 30
percent of the complainants noted dissatisfaction with the amount of
contact they had with investigators (see fig. 28). These results did not
vary significantly depending on whether cases had been investigated by
FHEO or FHAP agencies, but did differ according to the closure type. For
example, more than one-third of complainants whose cases were closed with
a finding of no-cause reported difficulties in contacting the
investigators, compared with 22 percent and 17 percent, respectively, of
complainants whose cases were closed administratively or conciliated.
Complainants with a no-cause finding were most dissatisfied with the
amount of contact they had with the investigator; 49 percent reported
dissatisfaction, compared with 18 percent and 14 percent, respectively,
for those having administrative or conciliation outcomes.

Figure 28: Complainant Views on Investigative Contact

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

Significant numbers of complainants reported that investigative staff
performed administrative functions acceptably and in a timely manner but
did only a fair or poor job on others. For example, complainants reported
that they were told their investigator's name about 83 percent of the
time, and 92 percent said they received their closure notifications. But
about 59 percent of complainants whose cases were administratively closed
reported that they were not told about any options they might have had for
pursuing a complaint. Further, overall many complainants believed that
staff did a fair to poor job of listening to them (nearly 40 percent),
explaining the investigative process (about 36 percent), investigating the
evidence (about 44 percent), interviewing their witnesses (nearly 40
percent), and asking for documents related to their cases (about 32
percent). These percentages were generally similar whether complainants'
cases had been handled by FHEO or a FHAP agency, with one exception:
Complainants indicated that FHAP agencies were slightly better at
interviewing witnesses (33 percent excellent or good) than FHEO (23
percent). We found more differences among complainants' experiences based
on their closure outcomes. Complainants with no-cause outcomes were more
likely to perceive difficulties in a variety of areas than complainants
with administrative closures or conciliated cases. Specifically,
complainants with no-cause outcomes

o more frequently reported problems receiving case updates (62 percent of
the time, compared with 34 percent for administrative closures and 23
percent for conciliations);

o were more likely to believe that investigators did a poor job of
investigating the evidence (68 percent of the time, compared with 37
percent for administrative closures and 20 percent for conciliations); and

o were more likely to believe that investigators did a poor job of
interviewing witnesses (58 percent of the time, compared with 35 percent
for administrative closures and 14 percent for conciliations).

Despite these views, most complainants reported that staff moved quickly
in conducting the investigation. About three-fifths of complainants said
that both FHEO and FHAP investigators were prompt in contacting them to
start an investigation, responding to questions, and completing an
investigation. However, complainants with no-cause outcomes typically
reported problems with timeliness at twice the rate of complainants with
different outcomes (see fig. 29).

Figure 29: Complainant Views on Timeliness for Certain Investigative
Activities, by Closure Type

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.

Finally, we found that complainants generally felt they were treated well
by investigators (see fig. 30), regardless of the type of agency that
investigated the complaint or the closure type, with one exception:
Complainants whose cases ended in a no-cause outcome usually felt less
well treated by staff. For example, an estimated 83 percent said that
investigators treated them with respect and courtesy, 74 percent said that
staff were interested in their complaint, about 72 percent believed that
staff were impartial, and around 71 percent found staff helpful. In
general, complainants had similar responses for FHEO and FHAP
investigations. As noted, those with a no-cause outcome were less
complimentary, and only 53 percent of them believed that the investigator
had been helpful, versus 78 percent and 90 percent, respectively, for
those with administrative or conciliation closures.

Figure 30: Complainant Views on Treatment by Investigative Staff

Note: Totals do not equal 100 percent because figure does not include
"don't know" responses. Brackets on each bar represent the sampling error
(confidence interval) for that estimate.

Conclusions

In our April 2004 report, we found that persons who have experienced
alleged discrimination in housing can sometimes face a lengthy wait before
their complaint is resolved. In preparing this current report, the results
of the test calls even though they are not generalizable to all potential
complainants and our survey of complainants that is generalizable- suggest
that some complainants may also face difficulties from the outset-that is,
during the intake phase-in contacting staff and presenting their initial
allegations. We also previously reported that without comprehensive,
reliable data on the dates when individuals make inquiries, FHEO cannot
judge how long complainants must wait before a FHAP agency undertakes an
investigation. For this report, our analysis of logged contacts indicates
that FHAP agencies and FHEO hubs did not enter into TEAPOTS all contacts
alleging Title VIII violations. Moreover, the discrepancies we observed
between the dates of logged initial contacts and the corresponding dates
entered into TEAPOTS as the beginning of the inquiries indicate that FHEO
does not have reliable data for measuring the extent to which its offices
and FHAP agencies meet the benchmark of 20 days for completing the intake
process.

Our review of case files and TEAPOTS data showed a lack of evidence that
investigations are consistently as thorough or expeditious as FHEO
guidance requires or recommends. HUD officials have noted their reliance
on TEAPOTS for assurance that cases are investigated in accordance with
applicable requirements and guidance. While the lack of evidence we
observed may be the result of not documenting certain actions, rather than
not carrying them out, the very lack of documentation or detailed TEAPOTS
records raises questions about HUD's ability to assure that investigations
are as thorough as they need to be.

The Fair Housing Act has from the outset mandated that persons alleging
housing discrimination be offered the opportunity to conciliate their
complaint with the other party. However, our review of case files and
TEAPOTS data showed a lack of evidence that conciliation attempts were
consistently made throughout the process, despite HUD's requirements that
these attempts be documented. The lack of documentation that conciliation
is offered raises questions about HUD's ability to assure that such
attempts are made as appropriate throughout the fair housing process.
Further, this lack of documentation, along with complainants' first-hand
experience and our observation during mock complaint calls that
conciliation was not discussed, suggest that FHEO and FHAP agencies are
not consistently offering conciliation, as required by the Act. We
recommended in our previous study that FHEO establish a way to identify
and share information on effective practices among its regional fair
housing offices and FHAP agencies. We observed during our previous study
and this study that some FHAP agencies have used independent mediators and
had staff not involved in a particular case attempt conciliation.
Officials who use these techniques point to their benefits in speeding up
the resolution of complaints, while offering the parties a satisfactory
outcome. In our last report, we concluded that FHEO's human capital
challenges serve to exacerbate the challenge of improving enforcement
practices. However, the identification and use of best practices may help
FHEO, as well as FHAP agencies, more effectively utilize their limited
resources.

Our work does not demonstrate that HUD failed to reach appropriate
decisions regarding any specific fair housing inquiry or investigation.
Further, our review of case files shows that many investigative
requirements were met, and former complainants we surveyed expressed
satisfaction with some aspects of their experience. Nevertheless, we
believe that our findings are cause for concern. Individuals who believe
they have experienced discrimination and make the effort to contact a fair
housing agency, but are unable to easily reach an intake staff person or
to expeditiously convey needed information, may simply give up and cease
cooperating. Further, our survey of former complainants shows that some
who successfully filed complaints have a sufficiently negative view of the
process that they would be unlikely to file a complaint again-even if they
were satisfied with the outcome of their case. Events of either type
diminish the Act's effectiveness in deterring acts of housing
discrimination or otherwise promoting fair housing practices.

Recommendations for Executive Action

To ensure that complainants are able to readily contact a fair housing
agency and file a complaint, we recommend that the HUD Secretary direct
the Assistant Secretary of FHEO to ensure that intake activities are
conducted consistently. Specific actions may include

o establishing clear standards for information that should be collected
with the initial contact;

o creating benchmarks and performance goals for the treatment of
complainants during the initial contact, including measures of
responsiveness, such as hold times and call-back timeliness, as well as
measures of completeness of initial information collection;

o developing special procedures for identifying and responding to
time-sensitive inquiries; and

o establishing means (including automation, where appropriate) of assuring
that standards, benchmarks, and special procedures are followed.

To improve the usefulness of TEAPOTS as a management control in assuring
that potential Title VIII-related contacts are identified and assessing
performance in meeting timeliness guidelines, we recommend that the HUD
Secretary direct the Assistant Secretary of FHEO to take the following two
actions:

o Specify that FHAP agencies use TEAPOTS for recording initial inquiry
dates for all inquiries, as defined in the Title VIII Handbook, that
allege housing discrimination.

o Require that the initial inquiry date reflect the first contact made by
the complainant, regardless of whether that contact was with FHEO or a
FHAP agency.

To enhance FHEO and FHAP agency ability to assess the thoroughness of
investigations, we recommend that the HUD Secretary direct the Assistant
Secretary of FHEO to take the following two actions:

o Establish documentation standards and appropriate controls to ensure
that required notifications of complaint, amendment, and closure are made
and received, and that 100-day letters are sent before an investigation
has reached 100 days.

o Clarify requirements for planning investigations, including specifying
when plans must be prepared, their content, and review and approval.

To ensure that some form of conciliation is made available for all
complainants, we recommend that the HUD Secretary direct the Assistant
Secretary of FHEO to take the following two actions:

o Work with FHAP agencies and others to develop best practices for
offering conciliation throughout the complaint process, including at its
outset.

o Ensure that investigators comply with requirements to document
conciliation attempts, and complainants' or respondents' declination of
conciliation assistance.

Agency Comments and Our Evaluation

We provided a draft of this report to HUD for its review and comment. We
received written comments from the department's General Deputy Assistant
Secretary for Fair Housing and Equal Opportunity. The letter, which is
included in appendix II, indicated a general agreement with our
conclusions and recommendations. The General Deputy Assistant Secretary
noted that FHEO conducts analyses of its programs and strives to
continually improve its operations and those of FHAP agencies in order to
ensure that complaints of housing discrimination are handled in an
effective and efficient manner. The letter also expressed confidence based
on the extensive internal reviews, final determinations, and requests for
reconsideration, in the integrity of the fair housing process, the
soundness of decisions and the competent professional service accorded
every party to the process. The letter noted a variety of initiatives that
have been implemented to improve the quality of investigations, including

o establishing the National Fair Housing Training Academy, which trains
fair housing professionals on fair housing law, critical thinking, and
interview techniques;

o completing revisions of its Intake, Investigation and Conciliation
sections of the Title VIII Handbook, which provides guidance to
investigators on case processing standards and sets nationwide policy;

o developing the FHEO-OGC Case processing Research Project, which focuses
on early interaction and continuous consultation between FHEO and OGC; and

o undergoing a business process re-engineering (BPR) to identify best
practices in the field among the FHAP agencies, as well as codifying
operations and procedures in headquarters.

The General Deputy Assistant Secretary commented that FHEO would, as
feasible, work to incorporate the recommendations into its policies and
procedures.

As agreed with your offices, unless you publicly release its contents
earlier, we plan no further distribution of this report until 30 days from
the report date. At that time, we will send copies to the Chair of the
Senate Committee on Banking, Housing and Urban Affairs; the Chair,
Subcommittee on Housing and Transportation, Committee on Banking, Housing
and Urban Affairs; the HUD Secretary; and other interested congressional
members and committees. We will also make copies available to others upon
request. In addition, this report will be available at no charge on GAO's
Web site at http://www.gao.gov.

If you or your staff have questions about this report, please contact me
at (202) 512-6878 or [email protected]. Contact points for our Offices of
Congressional Relations and Public Affairs may be found on the last page
of this report. GAO staff who made key contributions to this report are
listed in appendix IV.

David G. Wood Director, Financial Markets and Community Investment

Scope and MethodologyAppendix I

Our engagement scope was limited to fair housing investigations conducted
under Title VIII of the Civil Rights Act of 1968, as amended. We did not
address fair housing activities under Section 504 of the Rehabilitation
Act of 1973 or Title VI of the Civil Rights Act of 1964. For certain
analyses of the fair housing complaint process, we relied on national
samples of complaints closed during the last half of 2004-enabling us to
provide national estimates.

Intake

To determine the thoroughness of efforts by the Office of Fair Housing and
Equal Opportunity (FHEO) and Fair Housing Assistance Program (FHAP)
agencies during the intake process, we conducted two activities. First, we
asked intake staff to keep a log of the contacts they had with potential
complainants over a 4-week period and to note the proportion of contacts
alleging housing discrimination.

We designed an Intake Contact Log for staff to use that enabled us to
obtain never-before-collected data consistently across agencies for a set
time period. We asked the 10 FHEO offices, 36 state FHAP agencies, and 5
local FHAP agencies with the highest volume (based on number of complaints
filed during fiscal year 2004) to maintain the log over the 4-week period
from February 21 through March 21, 2005.1 All of the FHEO offices, local
FHAP agencies, and all but 4 of the state FHAP agencies agreed to maintain
our contact log. These offices represented 78 percent of the volume of
investigations in 2004. The log required intake staff to document

o the date each contact was received;

o the method of each contact (such as telephone, mail, e-mail);

o whether the contact involved a new potential complaint or a previously
existing complaint, or was a referral from another agency;

o whether the callers claimed that they had experienced housing
discrimination;

o whether the intake staff or supervisor believed the contact potentially
involved a jurisdictional Title VIII violation; and

o the name of the individual making the contact.2

Once we received the data, we reviewed them for consistency and logic.
Where we identified coding that was apparently not consistent with our
instruction, we called the staff that prepared the log for clarification.
In some cases, we needed to recode the log based on these conversations.

We focused our analysis on entries that the intake staff indicated
pertained to a potentially valid Title VIII issue, that included names,
and that they coded as new potential complaints. Using entries with names
allowed us to eliminate multiple contacts from the same person and report
statistics on people rather than on contacts. Once we identified the
unique names that met these criteria, we matched the names to records in a
TEAPOTS extract to determine how many of the new potential complaints
during the 4-week recording period were perfected. Specifically, we
requested an extract of the Department of Housing and Urban Development's
TEAPOTS database for the time period coinciding with the contact log
reporting period, plus 5 additional weeks to allow sufficient time for
perfecting complaints. We did not evaluate the judgments of the intake
staff in determining whether a contact should have been pursued. Since
FHAP agencies tended not to enter inquiries into TEAPOTS until the
complaint was ready to be perfected, the results of our name matching are
reported both in aggregate and separately for FHEO office and FHAP
agencies. We also reported on the total number of intake-related contacts
received during the reporting period, the proportion of contacts alleging
housing discrimination that the intake staff determined did not constitute
a valid Title VIII complaint, and percentages for each contact method
(telephone, e-mail, and walk-in).

In order to assess the degree to which intake staff obtained sufficient
and appropriate information to determine whether a contact should become a
Title VIII complaint, we designed a telephone "test call" program of
intake staff at the 10 HUD hubs and 36 state FHAP agencies. GAO analysts
posing as complainants contacted intake staff to file a mock complaint. We
excluded local FHAP agencies from telephone testing because these agencies
tend to have a low volume of investigations, compared with state FHAP
agencies. We placed one test call to each site using the same case
scenario with different identifying information, such as names and
addresses. Testers were trained to consistently volunteer only certain
information such as the "name" and description of what happened, and to
respond in standardized ways to questions asked by intake staff. The calls
were recorded and later coded against a list of information that might be
sought as part of the intake process. We developed this list of
information based on requirements and recommended practices derived from
multiple sources. Among these were requirements of the Fair Housing Act,
guidance provided by HUD in the form of policy and training manuals, and
training materials from the National Fair Housing Alliance, as well as
discussions with HUD and FHAP agency officials. We categorized the
information into four levels: information that (1) should always be
gathered at intake, (2) is potentially applicable to all complaints and
should be collected, (3) is relevant to a particular basis or protected
class, and (4) is considered by officials we spoke with and training
materials as a best practice.

We placed one pretest call to each of the 46 sites we planned to contact
to get a sense of how each location conducts intake. We adjusted the
design to account for differences in the intake process to the fullest
extent possible. For example, we allowed for scheduled intake interviews
for locations that only conducted intake calls on a scheduled basis. We
found that approximately 25 percent of the time we could expect to speak
with a person on initial contact who could complete the intake process. In
many cases, an agency staff member would perform an initial screening
before forwarding the call to the intake staff, and in other cases, the
call was forwarded to a voice mailbox. Based on our findings during the
pretest, we decided to evaluate not only the information collected by
intake staff, but also the number of attempts required to speak with a
live person, hold times, and the length of time that elapsed before staff
responded to voice-mail messages by returning calls. We included
information that callers volunteered and obvious items such as gender.
Because of the limitations of our sample-only one call to each site-the
results of our analysis are not generalizable to the entire population of
potential complainants in housing discrimination cases.

Investigation and Conciliation

To address the thoroughness of investigation procedures, including
conciliation attempts, we reviewed the documentation in 197 randomly
selected case files of housing discrimination cases completed during the
last 6 months of 2004 around the country (see table 1). We originally
sampled 205 cases, but we were unable to locate files or matching TEAPOTS
data records for 8 cases. The sample files included 58 cases closed
administratively, 63 cases that were conciliated without a finding of
reasonable cause, and 90 that were closed with a finding of no reasonable
cause. We oversampled administrative closures to ensure that we had a
sufficient number of files to permit estimates for this subpopulation. The
population of complaints and the sample we used are enumerated in tables 1
and 2. Because we followed a probability procedure based on random
selections, our sample is only one of a large number of samples that we
might have drawn. Since each sample could have provided different
estimates (sampling error), our results have confidence intervals of plus
or minus 8 percentage points or smaller, unless otherwise noted, at a 95
percent level of confidence. In other words, this interval would contain
the true value for the actual population for 95 percent of the samples we
could have drawn.

We also reviewed files for 12 of the 15 complaint investigations FHEO
concluded with a finding of "cause," and for which the adjudication
process, including any agency monitoring, had been completed during the
last 6 months of 2004. We could not locate files for the remaining cases.
All 12 files were identified by the Department of Justice as having met
these criteria. We did not review files for which a FHAP agency found
"cause" and that completed the adjudication because these could not be
identified.

Table 1: Location of FHEO and FHAP Agency Case Files Sampled

                                        

          FHEO region              FHEO file      FHAP agency file      Total 
Boston                                 16                     0         16 
New York                               17                     3         20 
Philadelphia                           17                     0         17 
Atlanta                                24                    10         34 
Chicago                                23                    11         34 
Fort Worth                              9                     8         17 
Kansas City, Kansas                    12                    14         26 
Denver                                  1                     4          5 
San Francisco                          23                     5         28 
Seattle                                 2                     6          8 
Total                                 144                    61        205 

Source: GAO analysis of FHEO data.

Table 2: Population and Sample Size of Complaints Closed from July 1,
2004, through December 31, 2004, by Organization and Closure Type

                                        

                     FHEO            FHAP             Total  
                                    agency                   
Closure type   Population Sample        Population Sample  Population Sample 
Administrative        208     19               335     33         543     52 
Conciliation          480     19             1,090     44       1,570     63 
No cause              563     23             1,657     67       2,220     90 
Total               1,251     61             3,082    144       4,333    205 

Source: GAO analysis of HUD data.

We examined the documentation in the files, as well as the full TEAPOTS
case summary for each case, to determine whether it demonstrated that the
investigator had met certain requirements and best practices for
conducting fair housing investigations. We identified these requirements
and best practices through reviewing the Fair Housing Act, implementing
regulations, FHEO's Title VIII Handbook and training material, and other
guidance. In addition, we interviewed FHEO officials at both HUD
headquarters and field offices in Atlanta, Chicago, and San Francisco. We
also interviewed FHAP agency officials in California, Georgia, Maryland,
South Carolina, and Virginia. We met with the National Fair Housing
Alliance and attended training at the John Marshall Law School on Fair
Housing enforcement and at HUD's National Fair Housing Training Academy.
We provided a draft summary of our criteria to FHEO officials and made
technical corrections based on their comments. To ensure the consistency
of our file review, we developed a structured data collection instrument.
We anchored each item on the instrument to the criteria identified above.
We pretested the instrument with several team members, and based upon this
test, modified the instrument to ensure clarity. For 10 percent of all
files reviewed, another team member reviewed the coding, to ensure its
accuracy.

Complainant Satisfaction

We surveyed a sample of complainants whose cases had been investigated and
closed by FHEO and FHAP agencies from July 1 to December 31, 2004, to
determine levels of satisfaction with the thoroughness, fairness,
timeliness, and outcomes of the intake and investigation process. We did
not include cases that proceeded to the adjudication process owing to a
finding of reasonable cause to avoid surveying complainants that may still
be involved in the adjudication process. The survey also provided
supplemental evidence for our analysis of the thoroughness of the intake
and investigation stages of the complaint process and the frequency of
conciliation.

We determined that there was a population of 4,327 fair housing complaints
(contact information was not provided for 6 of the original 4,333 cases)
that had ended in administrative closure, conciliation without
determination of cause, or a determination of no-cause for the 6-month
period between July 1, 2004, and December 31, 2004. The complainants of
record were mostly private individuals, but some were fair housing
organizations acting on their own behalf or on that of one or more
individuals. From a list obtained from HUD's TEAPOTS database, the PA
Consulting Group, a survey firm under contract to GAO, called the
complainants of record selected in the sample. The total sample of 1,675
was parceled out in seven individual waves over the field period, in an
attempt to use the smallest possible sample to achieve the quota of 575
completed interviews.  The sample was allocated across six categories-two
agency types by three closure types-to ensure that enough interviews were
conducted in each category to allow statistically valid comparisons
between them. (See table 4 for the distribution of the population, sample,
responses, and response rates across these categories.) With this
probability sample, each member of the population had a nonzero
probability of being included, and that probability could be computed for
any member. Each sampled complaint for which an interview was obtained was
subsequently weighted in the analysis to account statistically for all the
members of the population, including those who were not selected and those
who were selected but did not respond to the survey.

Beginning in early May 2005, GAO mailed letters notifying those sampled
complainants with valid mailing addresses of the upcoming survey and
encouraging them to participate. Calling to those complainants typically
began several days after the letters were mailed. The advance letters also
included an address correction form. Recipients were asked to revise any
incorrect information and return the form. A toll-free number also was
provided for recipients to ask any questions or to correct information
such as names and phone numbers. Calling began in early May and continued
for 7 weeks, ending on June 20, 2005.

For institutional complaints drawn into our sample, the interviewer helped
the organization's representative identify the specific complaint by
describing the issue, basis, and respondent name. Multiple interviews
could be conducted with the same institutional informant if more than one
complaint from that organization was randomly drawn. Thirty-three
interviews were completed on institutional complaints. Proxies (typically
family members, guardians, or other representatives) for the named
complainant were interviewed for 19 of the sampled complaints. Not all
sampled complaints met GAO's eligibility requirements for the survey: Some
complainants indicated that their cases were still open and experiencing
legal activity. GAO could not have any involvement with such complaints.
For the same reason, other complainants who said they had an ongoing
agreement with the other party resolving the complaint were also not
interviewed. Some complaint cases also became ineligible for the survey
due to the death of the complainant, a complainant's insistence that a
case was not a fair housing discrimination case, and complaints sampled
multiple times (more than four for institutional complainants and one for
individuals-see table 3).

Complainants with an issue or basis alleging racial discrimination because
the complainant was Hispanic received letters printed in both English and
Spanish. The survey was also administered by Spanish-speaking interviewers
when the complainant indicated a preference for speaking Spanish.

When named complainants could not be found using records provided by HUD,
interviewers used a variety of search techniques to try to locate
complainants, including calling alternate contacts named in the complaint
record and using directory assistance and online tracking services. For
example, if an address was available but no working phone number could be
found, interviewers used reverse directories and contacted neighbors by
phone to ask about the whereabouts of the named complainant. Once during
the fieldwork period, 547 noncontactable records were submitted to Lorton
Data's National Change of Address and Telephone Append services, resulting
in some updated phone numbers and mailing addresses. To maximize the
possibility of reaching complainants, multiple attempts were made over a
period of time on different days of the week and at different hours. The
number of attempts required to reach subjects ranged from 1 to 34, with an
average of 12.

Results from sample questionnaire surveys are subject to several types of
errors: failure of the sample frame to cover the study population,
measurement errors in administering the questionnaire, sampling errors,
nonresponse errors from failing to collect information from part of the
sample, and data processing error. To limit coverage errors, we used the
most recent available data from TEAPOTS to identify eligible complainants.
At the beginning of the interview, we also confirmed that the complainant
of record had lodged an actual fair housing discrimination complaint and
that the complaint had been closed as TEAPOTS indicated.

To limit measurement errors, we first took steps in the development of the
questionnaire to ensure that our questions gathered the information
intended. GAO asked knowledgeable representatives of fair housing
organizations and other government agencies to review early versions of
the instrument. We also conducted a pilot test of an early version of the
questionnaire with 26 complainants in December 2004 and seven pretests
with complainants in the study population during March 2005. Finally, the
telephone survey contractor completed nine pretests of the final
instrument in late April 2005. Interviewers were trained using materials
developed by GAO before the survey began and were routinely monitored
during interviews.

The survey is also subject to sampling error-our results have confidence
intervals of plus or minus 6 percentage points or smaller at a 95 percent
level of confidence.

Our survey received a low response rate, with only 38 percent of those
known or assumed to be eligible in our survey participating. If those who
did not respond might have answered our survey questions differently from
those who did, our estimates would be biased because we would have missed
the answers from a set of people with fundamentally different views. In
fact, response rates varied widely across the three different closure
types, which were associated with key variables in the survey such as
satisfaction with the complaint process and outcome. We tended to get
relatively more responses from complainants in conciliation cases than
from complainants in no-cause cases, and those whose cases were
conciliated tended to be more satisfied. However, we could address this
potential bias because we controlled the allocation of our sample across
the closure types. We could statistically adjust, or weight, responses by
closure type to bring them into proportion with the population and thus
account for the different nonresponse rate across those types, which
should compensate for the nonresponse bias. Nevertheless, the possibility
of bias in the results still remains. Our weighting adjustment only
compensates for differences in opinion between those with different
closure types and agency responsibility, not other characteristics that
may have been over- or under-represented in our responses and that may be
related to our survey questions.

To limit the possibility of data processing errors, the survey firm used a
computer-assisted telephone interviewing system that recorded electronic
data directly from the telephone interviewers' answers, and also checked
for missing data, inconsistencies, and unlikely answer patterns. Data
analysis programming was also independently verified.

Table 3: Distribution of the Population, Sample, Responses, and Response
Rates

                                        

               Organization         
                and closure         
                   type             
                               HUD                        FHAP          
                 Conciliation    No   Admin  Conciliation    No   Admin Total 
                              cause closure               cause closure 
Population             480   562     208         1,086 1,656     335 4,327 
Original               198   308     208           288   338     335 1,675 
sample                                                               
Ineligible              20    36      26            26    25      25   158 
Adjusted               178   272     182           262   313     310 1,517 
sample                                                               
Refused                  4    13       9            13    10       8    57 
Incapable                1     3       2             1     2       3    12 
Not                     75   157     109           139   187     206   873 
contacted                                                            
Complete                98    99      62           109   114      93   575 
interviews                                                           
Response               55%   36%     34%           42%   36%     30%   38% 
rate                                                                 

Source: GAO.

Comments from the Department of Housing and Urban DevelopmentAppendix II

GAO Contact and Staff AcknowledgmentsAppendix III

David G. Wood, (202) 512-6878

In addition to Mathew J. Scire (Assistant Director), Nicholas Alexander,
Carl Barden, Johnnie Barnes, Bernice Benta, Emily Chalmers, Arielle Cohen,
Paul Desaulniers, Grace Haskins, Robert Lowthian, Alexandra
Martin-Arseneau, Amanda Miller, Marc Molino, Jeff Pokras, Linda Rego, Carl
Ramirez, Beverly Ross, Paige Smith, Anita Visser, and Joan Vogel made key
contributions to this report.

(250204)

www.gao.gov/cgi-bin/getrpt?GAO-06-79.

To view the full product, including the scope

and methodology, click on the link above.

For more information, contact David G. Wood at (202) 512-6878 or
[email protected].

Highlights of GAO-06-79, a report to congressional requesters

October 2005

FAIR HOUSING

HUD Needs Better Assurance That Intake and Investigation Processes Are
Consistently Thorough

Each year, the Department of Housing and Urban Development's (HUD) Office
of Fair Housing and Equal Opportunity (FHEO) and related state and local
Fair Housing Assistance Program (FHAP) agencies receive and investigate
several thousand complaints of housing discrimination. These activities,
including required conciliation attempts, are directed by HUD's standards,
which are based on law, regulation, and best practices. GAO's 2004 report
examining trends in case outcomes raised questions about the quality and
consistency of the intake (the receipt of initial inquiries) and
investigation processes. This follow-up report assesses the thoroughness
of fair housing intake and investigation (including conciliation)
processes, and complainant satisfaction with the process.

What GAO Recommends

GAO makes recommendations to the HUD Secretary for improving the
thoroughness and timeliness of the fair housing complaint process,
including establishing standards and benchmarks for initial intake
activities, improving data needed to monitor timeliness of intake, and
improving planning and documentation of investigations.

HUD generally agreed with the report's conclusions and stated that it
would work to incorporate the recommendations into its policies and
procedures.

Evidence from several sources raises questions about the timeliness and
thoroughness of the intake process. Thirty percent of complainants GAO
surveyed noted that it was either somewhat or very difficult to reach a
live person the first time they contacted a fair housing agency. GAO
experienced similar difficulty in test calls it made to each of the 10
FHEO and 36 state FHAP agency intake centers. For example, 5 locations did
not respond to the test calls. Further, FHEO and FHAP agencies do not
consistently record in their automated information system contacts they
receive that they consider potential fair housing inquiries and timeliness
data are unreliable, limiting the system's effectiveness as a management
control.

GAO's review of a national random sample of 197 investigative case files
for investigations completed within the last 6 months of 2004 found
varying levels of documentation that FHEO and FHAP investigators met
investigative standards and followed recommended procedures. Further,
though the Fair Housing Act requires that agencies always attempt
conciliation to the extent feasible, only about a third of the files
showed evidence of such attempts. FHEO officials stated that the required
investigation and conciliation actions may have been taken but not
documented as required in case files.

According to GAO's survey of a national random sample of 575 complainants
whose complaint investigations were recently completed, about half were
either somewhat or very dissatisfied with the outcome of the fair housing
complaint process, and almost 40 percent would be unlikely to file a
complaint in the future. Although GAO and survey respondents found that
FHEO and FHAP agency staff were generally courteous and helpful, important
lapses remain in the complaint process that may affect not only how
complainants feel about the process but also how thoroughly and promptly
their cases are handled.

Survey Respondents' Overall Satisfaction with the Fair Housing Complaint
Process

Note: Totals do not equal 100 percent because figure does not include
"don't know" or "neither" responses. Brackets on each bar represent the
sampling error (confidence interval) for that estimate.
*** End of document. ***