Rural Housing Service: Updated Guidance and Additional Monitoring
Needed for Rental Assistance Distribution Process (13-SEP-04,	 
GAO-04-937).							 
                                                                 
The Rural Housing Service's (RHS) Section 521 rental assistance  
program provides rental subsidies for about 250,000 rural	 
tenants. Because the program has a waiting list of 80,000	 
eligible tenants and there are over 1.3 million rural low-income 
renters eligible for housing subsidies, it is important to	 
effectively distribute resources to extend assistance to as many 
needy tenants as possible. Therefore, GAO was asked to assess (1)
how RHS distributes rental assistance funds to properties and	 
tenants, (2) how RHS monitors the use of rental assistance funds,
and (3) whether there are sufficient internal controls in place  
to ensure that these funds are being effectively administered and
used.								 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-04-937 					        
    ACCNO:   A12335						        
  TITLE:     Rural Housing Service: Updated Guidance and Additional   
Monitoring Needed for Rental Assistance Distribution Process	 
     DATE:   09/13/2004 
  SUBJECT:   Disadvantaged persons				 
	     Funds management					 
	     Internal controls					 
	     Low income housing 				 
	     Monitoring 					 
	     Rent subsidies					 
	     Rental housing					 
	     Rental rates					 
	     Rural housing programs				 
	     RHA Automated Multi-Housing Accounting		 
	     System						 
                                                                 
	     RHS Multi-Family Housing Information		 
	     System						 
                                                                 
	     RHS Multifamily Direct Rural Rental		 
	     Housing Loans Program				 
                                                                 
	     RHS Multifamily Housing Farm Labor Loans		 
	     Program						 
                                                                 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-04-937

United States Government Accountability Office

 GAO 	Report to the Ranking Minority Member, Subcommittee on Agriculture, Rural
     Development, and Related Agencies, Senate Committee on Appropriations

September 2004

RURAL HOUSING SERVICE

    Updated Guidance and Additional Monitoring Needed for Rental Assistance
                              Distribution Process

                                       a

GAO-04-937 

Highlights of GAO-04-937, a report to Ranking Minority Member,
Subcommittee on Agriculture, Rural Development, and Related Agencies,
Senate Committee on Appropriations.

The Rural Housing Service's (RHS) Section 521 rental assistance program
provides rental subsidies for about 250,000 rural tenants. Because the
program has a waiting list of 80,000 eligible tenants and there are over
1.3 million rural lowincome renters eligible for housing subsidies, it is
important to effectively distribute resources to extend assistance to as
many needy tenants as possible. Therefore, GAO was asked to assess (1) how
RHS distributes rental assistance funds to properties and tenants, (2) how
RHS monitors the use of rental assistance funds, and (3) whether there are
sufficient internal controls in place to ensure that these funds are being
effectively administered and used.

GAO recommends that the Secretary of Agriculture require that program
officials establish centralized guidance on transferring unused rental
assistance, improve sampling methods to select tenant household files for
supervisory reviews, and improve verification of tenant information. To
help the agency verify tenant information, Congress may wish to consider
giving RHS access to the Department of Health and Human Services' National
Directory of New Hires.

The Department of Agriculture generally agreed with GAO's recommendations.

www.gao.gov/cgi-bin/getrpt?GAO-04-937.

To view the full product, including the scope and methodology, click on
the link above. To view the survey results, click on the following link
www.gao.gov/cgi-bin/getrpt?GAO-04978SP. For more information, contact
William B. Shear at (202) 512-4325 or [email protected].

September 2004

RURAL HOUSING SERVICE

Updated Guidance and Additional Monitoring Needed for Rental Assistance
Distribution Process

RHS's national office pays rental assistance funds to property owners
through a variety of contracts and transfers of unused rental assistance
from one property to another. RHS staff calculate the payment amounts
using information provided by tenants and verified by property owners. The
current system requires RHS staff to enter payment data into the agency's
database manually. However, RHS's national office is upgrading two systems
to streamline the payment process. But some local RHS staff are concerned
that the new process may eliminate a layer of review.

RHS's national, state, and local offices share responsibility for
monitoring the rental assistance program, with the local offices
performing the primary review every 3 years. The national office outlines
the monitoring activities for the rental assistance program in its Rural
Development Instructions and provides other guidance, such as
administrative notices and unnumbered letters. We found that state and
local offices follow this guidance inconsistently and sometimes consider
it unclear. National office staff are planning a number of initiatives to
increase monitoring of the rental assistance program, but a number of key
management positions in the national office are unfilled, which could
limit the effectiveness of program monitoring.

RHS's internal controls do not provide adequate oversight of rental
assistance funds because RHS guidance is inadequate and tenants' incomes
are not adequately verified. First, insufficient guidance on the transfer
process limits RHS's ability to move unused rental assistance to
properties that have tenants with the greatest unmet need. Second, the
triennial supervisory review does not provide reasonable assurance that
tenants income and assets, and ultimately rental assistance payments, are
adequately verified. Alternate methods of verifying tenant information
also have limited effectiveness but could help improve internal control if
properly designed and implemented.

How RHS Uses Information from Tenants to Determine Rental Subsidies

Source: GAO.

Contents

  Letter

Results in Brief
Background
RHS Distributes Rental Assistance Funds through Contracts and

Sets Per-Unit Payments Based on Tenants' Verified Income Levels

RHS's National and State Offices Monitor the Use Of Rental Assistance,
with Local Offices Performing the Primary Review Every 3 Years

RHS's Internal Controls Do Not Provide Adequate Oversight of

Rental Assistance Funds Conclusions Recommendations for Executive Action
Matter for Congressional Consideration Agency Comments and Our Evaluation

1 2 4

8

14

21 32 33 33 33

Appendixes

Appendix I: Appendix II:

Appendix III:

Appendix IV: Appendix V: Scope and Methodology

State Office Survey Responses for Rural Rental Housing (RRH) Properties

State Office Survey Responses for Off Farm Labor Housing (FLH) Properties

Comments from the U.S. Department of Agriculture

GAO Contacts and Staff Acknowledgments

GAO Contacts
Staff Acknowledgments

36

39

41

42

45 45 45

Table Table 1:	Triennial Supervisory Review Workload, Selected RHS Local
Offices

                                Duties of RHS National, State, and Local      
Figures Figure 1: Figure 2:   Offices Rental Assistance Distribution   6 9
                                                Process                   
                     Figure 3:  How RHS Uses Information from Tenants to  
                                               Determine                  
                                            Rental Subsidies               11 
                     Figure 4: Centralized and Decentralized Processing,   13 
                                                by State                  
                     Figure 5:   Key Management Positions in the Rural    
                                            Housing Service,              
                                            as of July 2004                20 

Contents

Figure 6: Frequency of Supervisory Reviews of Tenant Household Files 25
Figure 7: Process Used to Verify the Accuracy of Rental Assistance
Payments 29 Figure 8: States with Wage-Matching Agreements 31

Abbreviations

AMAS Automated Multi-Housing Accounting System
FLH Farm Labor Housing
HUD U.S. Department of Housing and Urban Development
MFIS Multi-Family Housing Information System
MINC Management Agency Interactive Network Connection
RD Rural Development
RHS Rural Housing Service
RRH Rural Rental Housing
USDA U.S. Department of Agriculture

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

A

United States Government Accountability Office Washington, D.C. 20548

September 13, 2004

The Honorable Herbert Kohl

Ranking Minority Member

Subcommittee on Agriculture, Rural Development and Related Agencies
Committee on Appropriations United States Senate

Dear Senator Kohl:

The Rural Housing Service (RHS) of the U. S. Department of Agriculture
(USDA) provides rental subsidies through the Section 521 rental assistance
program to about 250,000 low-income rural tenants living in federally
subsidized properties.1 The rental subsidies, which are provided to
property owners through multiyear contracts, are intended to limit
tenants' rent payments to 30 percent of the household's adjusted monthly
income. With a fiscal year 2004 budget of almost $600 million, the program
is RHS's largest line-item appropriation. Given its size, internal
controls are a significant factor in administering the program in order to
distribute resources effectively and reduce the possibilities for fraud,
waste, and abuse. Because the program already has a waiting list of 80,000
eligible tenants and over 1.3 million low-income renters not receiving
housing subsidies reside throughout rural America, effectively
distributing resources is important in extending assistance to as many
needy tenants as possible.2

We recently reported on the activity level of rental assistance contracts
and the accuracy of RHS's budget estimates for the Section 521 rental

1The Section 515 Multifamily Direct Rural Rental Housing Loans program
provides subsidized loans to developers to build rental housing. The
Section 514 Multifamily Housing Farm Labor Loans program provides grants
and subsidized loans to build rental housing for farm laborers, either off
or on the farmer's property. Only "off-farm" labor housing is eligible for
rental assistance subsidies. All properties must be established on a
nonprofit or limited profit basis to receive rental assistance.

2GAO analysis of the 2003 American Housing Survey. This figure excludes
cash renters, and rural areas include rural parts of metropolitan areas
and excludes urbanized parts of nonmetropolitan areas. The American
Housing Survey is the largest, regular national housing sample survey in
the United States. The U.S. Census Bureau conducts the American Housing
Survey to obtain up-to-date housing statistics for the Department of
Housing and Urban Development.

assistance program.3 To further assist you in your oversight of the
program, you also asked us to assess (1) how RHS distributes rental
assistance funds to properties and tenants, (2) how RHS monitors the use
of rental assistance funds, and (3) whether RHS has sufficient internal
controls in place to ensure that these funds are being effectively
administered and used.

To respond to these objectives, we collected written information from RHS
and talked with its representatives about the processes for distributing
rental assistance and the means of monitoring those processes. We reviewed
RHS's guidance for the program and internal control standards that apply
to the processes. We also conducted a nationwide survey of all RHS state
and local offices with rental assistance responsibilities. Finally, we
visited RHS state and local offices and property owners or managers in
five states to obtain documentation related to specific aspects of the
rental assistance distribution process and to observe how RHS staff
implement the internal controls that RHS has established to oversee the
program. Details of our scope and methodology appear in appendix I, and
appendixes II and III contain state office responses to a number of survey
items. The survey and a more complete tabulation of the results can be
viewed at www.gao.gov/cgi-bin/getrpt?GAO-04-978SP.

We conducted our work from September 2003 through June 2004 in a number of
locations, including Connecticut, Massachusetts, Mississippi, New York,
South Carolina, Utah, and Washington, D.C., in accordance with generally
accepted government auditing standards.

Results in Brief	Rental assistance funds are paid monthly to property
owners through multiyear, renewable contracts that designate a set amount
of funding per rental unit in RHS-subsidized properties. Property owners
may also request that RHS transfer unused rental assistance from one
property to another. The actual amount of the payments property owners
receive is calculated using information provided by tenants, verified by
property owners, and reviewed by staff at RHS state and local offices.
Plans to streamline the payment process are under way, but some RHS staff
are concerned that the new process may eliminate a layer of review.

3Rural Housing Service: Standardization of Budget Estimation Processes
Needed for Rental Assistance Program, GAO-04-424, (Washington, D.C.: March
25, 2004).

Responsibility for monitoring the rental assistance program is divided
among RHS's national, state, and local offices, with the local offices
performing the primary review of the rental assistance distribution
process. The national office provides guidance on procedures for
monitoring the distribution of rental assistance funds, but we found that
state and local offices apply the guidance inconsistently and that some of
these offices consider it unclear. Following the guidance, local offices
conduct supervisory reviews every 3 years, and these reviews are RHS's
primary tool for detecting misreporting of tenant income, and ultimately,
unauthorized rental assistance payments. Initiatives to increase
monitoring of the rental assistance program are under way at the national
office. However, unfilled management positions at the national level could
limit RHS's ability to effectively monitor the program.

RHS's internal controls do not provide adequate oversight of rental
assistance funds. First, insufficient and unclear guidance on the transfer
process limits RHS's ability to transfer unused rental assistance to
properties with the most needy tenants. Second, RHS's triennial
supervisory reviews, the primary internal control activity to verify the
information tenants and property owners submit, are conducted infrequently
using samples that may be too small to effectively mitigate the risks
associated with providing unauthorized rental assistance. RHS uses
alternate methods, such as internal database checks and wage matching, to
verify tenant and payment information. These methods have limited
effectiveness in governing the rental assistance distribution process but
could help to improve internal control if they were properly designed and
implemented.

This report contains three recommendations for executive action and one
matter for congressional consideration.

o 	To ensure that rental assistance funds are effectively distributed to
properties that have tenants with the greatest need in the multifamily
housing portfolio, we recommend that the Secretary of Agriculture require
program officials to (1) establish centralized guidance on transferring
unused rental assistance, (2) improve sampling methods to ensure that a
sufficient number of tenant household files are selected for supervisory
reviews, and (3) improve verification of tenant information, including
more effective use of alternate methods of income verification.

o 	 To enable RHS to improve its verification of tenant income, this
report includes a matter for congressional consideration. We recommend
that RHS be added to Section 453 (j)(7) of the Social Security Act,
"Information comparisons for housing assistance programs."4 This action
would give RHS access to the National Directory of New Hires administered
by the Department of Health and Human Services.

In response to our recommendations, RHS agreed to provide centralized
guidance on transferring unused rental assistance through a new
administrative notice and to institute broader use of alternate income
verification methods, such as improved access to wage matching and new
hires data. However, RHS did not agree to change its sampling methods for
selecting tenant files when performing supervisory reviews. Instead, RHS
said that it would revamp the supervisory review process to improve its
effectiveness and consistency. We agree that the supervisory review
process should be revised to improve its effectiveness and consistency and
are encouraged that USDA is taking steps to accomplish this. However, to
make this process an effective control to provide reasonable assurance
that rental assistance payments are correct, RHS must also improve its
sampling methodology, using a statistically valid sample to estimate an
overall error rate. Thereby, RHS would have a firmer basis for determining
an approximate sample size for its reviews.

Background	The Section 521 rental assistance program, started in 1978, is
administered by RHS's Multifamily Housing Portfolio Management Division.
The program provides rental assistance for tenants living in units created
through RHS's Multifamily Direct Rural Rental Housing Loans and
Multifamily Housing Farm Labor Loans programs. These programs provide
loans subsidized with interest rates as low as 1 percent to help build
rental housing for rural residents and farm workers. Under the rental
assistance program, eligible tenants pay no more than 30 percent of their
income toward the rent, and RHS pays the balance to the property owner.5

442 USC 653 (j)(7).

5Eligible tenants are persons with very low and low incomes, the elderly,
and persons with disabilities who are unable to pay the basic monthly rent
within 30 percent of their adjusted monthly income. Very low income is
defined as below 50 percent of the area median income; low income is
between 50 and 80 percent of area median income.

RHS allocates rental assistance subsidies through 5-year contracts with
property owners; 20-year contracts were also issued to units in newly
constructed properties from 1978 through 1982. In fiscal year 2004,
contracts were issued for a term of 4 years. The contracts specify that
owners will receive payments on behalf of tenants in a designated number
of units at the property. Contracts may be renewed as many times as funds
are made available, and additional rental assistance units may be
allocated if funds become available.

Property owners are required to follow RHS guidance in managing their
properties.6 Each month, they must document how many units designated for
rental assistance are occupied. If a unit is empty and rental assistance
is not being used, the property owner is required to assign another
eligible tenant. If no eligible tenants are waiting for the unit, the
property owner can work with RHS to transfer the rental assistance to
another property.

As shown in figure 1, RHS's national, state, and local offices share
administrative duties for the rental assistance program.

6Rural Development Instructions allow property owners to delegate or
assign management responsibilities, in full or in part, to a property
manager. Therefore, the term "property owner" used throughout this report
may also refer to a designated property manager, unless property manager
is specifically stated.

The national office also distributes administrative notices and unnumbered
letters on an as-needed basis. Administrative notices clarify for state
and local offices activities outlined in RD Instructions. For example, the
national office plans to clarify the tenant income verification process
and provide further guidance on monitoring unused rental assistance for
transfer through forthcoming administrative notices. Unnumbered letters,
which address specific issues and are less formal than administrative
notices, relay important information to state and local offices, but they
do not pertain to policies and procedures. For example, an unnumbered
letter released in December 2003 included data aggregated at the national
office broken down by state and by rental assistance distribution goals,
such as percentage of supervisory reviews completed, unit vacancies, and
unused rental assistance.

The national office also estimates program budgets; allocates funds; and
tracks nationwide program statistics. State and local offices work
directly with property owners, property management companies, and tenants
to monitor the program. State and local offices are responsible for
conducting financial, management, and physical reviews of the properties;
executing rental assistance contracts with property owners; approving rent
increases; and processing rental assistance payments. State and local
office staff also collect and maintain property and tenant data for their
areas.

Support for the program is also provided through two offices in St. Louis,
Missouri. The Rural Housing Service Branch of the Information Resources
Management Office maintains the database systems used to manage program
data. The Office of the Deputy Chief Financial Officer uses the program
data to generate and maintain the general ledger and financial statements
for the program. RHS uses two key database systems to manage their
multifamily housing programs, including the rental assistance program. The
Automated Multi-Housing Accounting System (AMAS) stores data on rental
assistance contract obligations, rental assistance payments to property
owners, and transfers of rental assistance. The Multi-Family Housing
Information System (MFIS) tracks property and tenant information that
determine the monthly request for rental assistance.

As mandated by the Federal Managers' Financial Integrity Act of 1982, the
Comptroller General issues standards for internal control in the federal

government.7 These standards provide the overall framework for
establishing and maintaining internal control and for identifying and
addressing major performance and management challenges and areas at
greatest risk of fraud, waste, abuse, and mismanagement. According to
these standards, internal control comprises the plans, methods, and
procedures used to meet missions, goals, and objectives. It is the first
line of defense in safeguarding assets, and preventing and detecting fraud
and errors. Internal control, which is synonymous with management control,
helps government program managers achieve desired results through
effective stewardship of public resources. No matter how well designed and
operated, internal control cannot provide absolute assurance that all
agency objectives will be met, and thus, once in place, internal control
provides reasonable, not absolute, assurance of meeting agency objectives.

RHS Distributes Rental Assistance Funds through Contracts and Sets
Per-Unit Payments Based on Tenants' Verified Income Levels

RHS's national office provides rental assistance funds to property owners
through a variety of contracts. Rental assistance can also be transferred,
under certain conditions, from one property to another. RHS calculates the
payments property owners receive using information supplied by tenants and
verified by property owners, who then use the information to create a
tenant certification form. Tenants' income must be certified yearly on the
tenant certification form or whenever significant changes occur-for
instance, when income increases by a specified amount. Property owners
send the tenant certification form to the responsible state or local
office, where RHS staff review the information and calculate the amount of
the subsidy. In order to receive the monthly payments, property owners
must provide the appropriate RHS office with a project worksheet-a monthly
statement showing how many units that are eligible for subsidies are
occupied by tenants in the program. Figure 2 shows the steps that are
involved in the distribution process. Plans are under way to streamline
the payment process, although local office staff are concerned this may
eliminate their chance to review the information on the project worksheet.

7Internal Control: Standards for Internal Control in the Federal
Government, GAO, (Washington, D.C.: November 1999).

Figure 2: Rental Assistance Distribution Process

                                  Source: GAO.

Funds Are Distributed to Properties through Contracts and Transfers of
Unused Rental Assistance

The rental assistance program provides four types of contracts: renewal,
construction, servicing, and preservation. RHS officials report that in
the last few years, 96 percent of annual rental assistance appropriations
have been used for renewal contracts-that is, contracts that renew
expiring obligations at existing properties. Construction contracts, which
support new units in recently constructed properties, account for only 1
percent. The remaining 3 percent cover servicing and preservation
contracts. Property owners who take out loans to rehabilitate properties
and then raise rents to cover the servicing costs associated with the
loan, for example, are eligible for servicing contracts that provide
additional rental assistance to help tenants cope with the rent increase.
Preservation contracts are offered as an incentive to property owners who
are considering leaving the program and provide subsidies for additional
units on the property.

According to the authorizing statute, unused rental assistance is to be
transferred to properties where the assistance is needed. Transfers can
occur when state and local offices identify unused rental assistance at a
property or when a property owner requests such a transfer. For example, a
property owner may want to transfer unused rental assistance from a
property that has no eligible potential tenants to one that has many.
During fiscal year 2003 unused rental assistance on approximately 6,800
units, or about 3 percent of all rental assistance units-with a total
value of $67 million-were transferred through about 2,000 transactions. An
RHS national office official noted that a portion of these units were
rental assistance transfers accompanying a property transfer, during which
an eligible property owner assumes another property owner's loan.

Property owners may initiate the transfer of unused rental assistance from
their properties at any time. If the property owner has not used the
rental assistance on one or more of the covered units for any consecutive
12month period, RHS may transfer the unused rental assistance, as long as
at least one unit at the property remains covered.8 In either case, before
the transfer can occur, RHS must verify that the rental assistance has
been properly marketed by the owner and is not needed at the current
property. Property owners may appeal decisions to transfer rental
assistance from their properties by contacting USDA's National Appeals
Division, an independent office within USDA.

RHS Makes Payments on the Basis of Information Provided by Tenants and
Verified by Property Owners

As figure 2 shows, a number of steps take place between the time rental
assistance contracts are made or transferred and the time property owners
receive payments. The two key steps in this process involve certifying
tenants' income information on the tenant certification form and using
that information to calculate the payments to property owners on the
project worksheet. (fig. 3).

8The consecutive 12-month period can begin only after the end of the first
year of the rental assistance agreement. RHS's proposed rule to update the
current RD Instructions gives the agency the authority to transfer rental
assistance that has not been used for a 4-month period. (See 7 C.F.R. Part
3560.259, Proposed Rule).

Figure 3: How RHS Uses Information from Tenants to Determine Rental
Subsidies

Source: GAO.

RD Instructions require that tenants be certified at least annually so
that state and local office staff can document their income and determine
their rental contributions. Property owners certify tenants initially and
then every year, or when a tenant's income rises by $40 or more per month
($480 per year) or decreases by $20 or more per month ($240 per year). In
these cases, the income must be verified again. According to RD
Instructions, the property owner must obtain verification of the tenant's
stated income, in writing, from the tenant's employer. If the owner is not
able to immediately obtain written verification, the tenant's income may
be temporarily verified by examining income checks or check stubs. Income
listed from other sources, such as Social Security or public assistance,
must be verified in writing by those sources. Although tenant
certifications and income verifications help ensure accurate payments to
property owners, errors can occur when tenants do not accurately report
all sources of income or when property owners do not adequately verify
them.

Calculating the amount of rental assistance and submitting the information
to RHS is the second key step. Project worksheets document the rent and
income levels of tenants for whom the property owner can request rental

assistance and calculate the amount of rental assistance due.9 Property
owners submit the project worksheets by mail to state or local offices,
where, according to RHS guidance, staff are to compare the worksheet
information with information from MFIS. Once a month, MFIS also generates
project worksheets, based on the previous month's occupancy reports and
the tenant certifications in the system. RHS staff finish processing the
worksheets by entering the information into AMAS, triggering payments to
the property owners.

In over half of the states, individual local offices are responsible for
processing project worksheets. However, approximately 39 percent (17 out
of 44) of the state offices we surveyed have chosen to centralize their
processing function (fig. 4).10

9Rental assistance is used in conjunction with RHS loans, and is provided
directly to the owners of some RHS-financed properties. The amount of
rental assistance requested by the property owner is subtracted from the
amount of the property owner's monthly mortgage payment. If the expected
amount of rental assistance exceeds the amount of the property owner's
monthly mortgage payment, then RHS will issue a rental assistance check to
the property owner.

10We surveyed only RHS state offices with rental assistance
responsibilities. Also, in certain instances RHS has a single state office
for selected groupings of states. For example, the state office for
Connecticut, Massachusetts, and Rhode Island is located in Massachusetts.
Thus, the total number of state offices is less than 50.

Figure 4: Centralized and Decentralized Processing, by State

Source: GAO.

Note: In certain instances, RHS has a single state office for selected
groupings of states. To depict centralized status in these selected state
groupings, this figure only shades the state where the state office is
located. See appendix I for more details on RHS's state groupings.

Program officials in one of these states told us their centralized
processing center promotes consistency in administering regulatory
requirements. In another centralized state, an official stated that their
centralized payment center had helped to cut costs and increase
satisfaction, because property owners had one place to call when questions
or payment issues arose. However, program officials in two decentralized
states told us that they believed that centralized payment systems
distanced RHS staff from the properties and property owners or managers.

RHS's Plans to Streamline the Payment Process Have Raised Concerns about
Eliminating a Layer of Review

RHS's national office is upgrading two systems to streamline the payment
process. The Management Agency Interactive Network Connection (MINC) is a
Web-based electronic transmission process that will upgrade the current
modem-based process, known as the Industry Interface, that property owners
can use to transmit tenant certifications to RHS. RHS will also be
upgrading MFIS in order to create electronic project worksheets, so that
state and local offices will no longer be required to enter data from the
project worksheets manually into AMAS. Once the property owners approve
their electronic project worksheets through MINC, the information will be
transmitted automatically into AMAS to trigger rental assistance payments.
As of May 2004, five states have been using the MFIS upgrade in
conjunction with MINC to process payments and the national office expects
all states will be using it by the end of calendar year 2004.

Local office staff whom we spoke with during one of our site visits were
concerned that allowing property owners to approve and submit project
worksheets electronically will eliminate the local office's chance to
review the information while entering it into AMAS. National office staff
overseeing these changes stated that electronic project worksheets would
not affect this review, because the property owner would not be able to
make any changes directly to the electronic project worksheet if he or she
did not agree with the amount of rental assistance calculated. Rather, any
modifications to the project worksheets would have to be submitted
separately to RHS.

RHS's National and State Offices Monitor the Use Of Rental Assistance,
with Local Offices Performing the Primary Review Every 3 Years

RD Instructions provide guidance for monitoring the use of rental
assistance funds. The guidance requires reviews at the national, state,
and local levels. However, the triennial supervisory review is the primary
tool for verifying the accuracy of rental assistance payments. National
office staff are planning a number of initiatives to increase rental
assistance monitoring, but a number of key management positions in the
national office have been unfilled, potentially limiting the effectiveness
of monitoring efforts.

Guidance from the National Office Establishes Monitoring Procedures but Is
Used Inconsistently and Considered Unclear

Two main sets of RD Instructions establish procedures for distributing
rental assistance. These include:

o 	RD Instruction 1930-C, Management and Supervision of Multiple Family
Housing Borrowers and Grant Recipients, particularly exhibits B and E,
which establishes criteria for tenant and property eligibility for rental
assistance, among other things.

o 	RD Instruction 1951-K, Exhibit A, Processing Multiple Housing Payments,
which outlines the rental assistance payment process.

Although four offices we visited reported following RD Instruction 1930-C
to process tenant certifications, we found little consistency in the
guidance RHS offices used to process payments, specifically when
processing project worksheets. Three offices we visited reported using the
AMAS Online Cash Manual, which provides instructions on using the AMAS
database to process payments, but two of those offices also reported using
different parts of the RD Instructions issued by the national office. One
of these two offices reported using the RD Instruction 1951-K. One other
office was unaware of official guidance, and another office told us that
they referred questions to their AMAS coordinator in the state office.

Likewise, two sets of RD Instructions provide explicit procedures for
monitoring the rental assistance program, including the purpose,
frequency, content, and planning and follow-up procedures:

o 	RD Instruction 2006-M, Management Control System, which outlines the
reviews used at the national and state office levels: the Management
Control Review and the State Internal Review; and

o 	RD Instruction 1930-C, which outlines the triennial supervisory review
for local offices.

The national office conducts the Management Control Review on a 5-year
cycle to review rental assistance administration in randomly selected
states. A Management Control Advisory Group, which is made up of national
and local office employees, determines the time frames, objectives, and
work plan for the Management Control Review cycle. Generally, the
objectives are designed to assess state-level administrative operations,
including financial management, internal control, and automated systems.
RHS national office staff visit the selected properties and review a
sample of tenant household files from each. Based on the

results of the review, the national office has the authority to make
recommendations to the state offices for improving rental assistance
administration.

State offices are required to conduct the State Internal Review at least
once every 5 years. This review is designed to, among other things, ensure
that RHS local offices properly implement Rural Development policies and
procedures. During these reviews, state office staff evaluate program and
administrative functions at local offices. Reports on the findings from
these reviews are provided to the state director.

Local offices are to conduct supervisory reviews at least every 3 years,
examining a random sample of tenant household files at each property in
their portfolios, and RD Instructions mandate more frequent reviews for
delinquent or problem properties. Local office staff in one state we
visited told us that they conduct approximately one supervisory review per
month in order to review each property once every 3 years.11 To conduct
the review, staff are to pull a random sample of at least six tenant
household files at each property, or review all tenant household files if
the property has six units or fewer. RHS provides sample forms including a
pre-visit worksheet to help staff sample units, plan objectives, and
assemble required documentation for each visit.12 Table 1 shows the number
of rental assistance units subject to triennial supervisory review at each
of the five local offices we visited.

11Based on a Connecticut local office portfolio of 34 properties and 920
rental assistance units.

12RD Instruction 1930-C, Exhibits F and F-1.

Table 1: Triennial Supervisory Review Workload, Selected RHS Local Offices

                                                             Number of rental 
        State where local Number of rental         Number of       assistance 
                                                   dedicated 
        office is located assistance units      office staff  units per staff 
              Connecticut              920                 2 
              Mississippi              500                 4 
                 New York              480                 2 
           South Carolina            2,149                 2            1,075 
                     Utah              368                 2 

Source: GAO.

In written responses to survey questions, certain respondents cited issues
related to RHS guidance in the conduct of their rental assistance
responsibilities. For example, one survey respondent stated that RHS
guidance needed to be simplified and issued in a timely manner. Another
survey respondent stated that RHS guidance is sometimes unclear. This
respondent added that administrative notices helped to some degree but
added that the office was receiving more guidance from unnumbered letters,
which do not carry the same force as RD Instructions or administrative
notices.

Despite these comments, 80 percent of respondents considered national
administrative notices helpful or very helpful at clarifying policies and
procedures established by the national office. However, approximately 50
percent of respondents were unable to say how many notices the national
office had issued between January 2003 and January 2004.

Triennial Supervisory Reviews Are RHS's Primary Monitoring Tool for the
Rental Assistance Program

The triennial supervisory review is the chief activity RHS local offices
use to verify tenants' income, assets, and expenses-that is, the
information on which the tenant certifications and project worksheets, and
ultimately rental assistance payments, are based. RHS staff look for
supporting documentation that confirms tenants' eligibility and income
during tenant household file reviews and examine the property owner or
manager's waiting list, rental rates, and budget. While property owners
are required to collect supporting documentation for tenants' claims of
income, assets, and expenses -a process referred to as third party income
verification-RHS staff do not see the original documentation tenants
provide at any time

other than during triennial supervisory reviews.13 Local office staff in
all five states selected for site visits told us that triennial
supervisory reviews were the mechanism they used to monitor the accuracy
and completeness of tenant certification data. Further information
collected from the local office staff emphasized the value of triennial
supervisory reviews. For example, four out of five local offices we
visited recalled finding errors or questionable tenant documentation
during triennial supervisory reviews.

RHS uses the triennial supervisory review and other mechanisms to detect
unauthorized rental assistance payments but faces challenges recapturing
the unauthorized payments. Tenant household file reviews can help detect
erroneous claims that lead to unauthorized payments. For example, during a
triennial supervisory review, RHS staff in one state found that a property
owner had allowed a tenant to overstate medical expenses, which could
lower the tenant's net income and thus increase the payment RHS makes on
the rental unit. Employers may also file erroneous third-party income
verification forms. According to a property owner we visited, such errors
can occur when employers falsify income information in order to maximize
rental assistance payments for their employees. RHS can detect such errors
through wage matching, the process of comparing statements of income by
tenants or employers to data collected by state wage information
collection agencies.

According to RHS officials, recapturing unauthorized rental assistance
payments once they have been made is difficult. Though RHS can report
false statements of income to state agencies, these agencies seldom have
recourse against tenants or employers. When surveyed, 49 percent of RHS
staff stated that they collect unauthorized rental assistance payments
from property owners. Owners can, in turn, attempt to collect the
unauthorized payments from tenants through a repayment agreement.
Twenty-nine percent of the RHS offices that indicated that their office is
able to reclaim unauthorized rental assistance reported that they had
attempted to initiate payment agreements with tenants. However, in written
responses to survey questions, one respondent noted that the office is not
able to recapture unauthorized rental assistance in over 95 percent of the
cases. Another survey respondent stated that the office is successful in
recapturing unauthorized rental assistance in about 30 to 50 percent of
the cases.

13In one local office we visited, RHS staff stated that they have
augmented their review process to include spot checks of tenant
documentation on an annual basis.

Additionally, survey respondents stated that tenants typically vacated the
units before the debt was repaid.

Initiatives to Increase Monitoring of the Rental Assistance Program Could
Be Hindered by the Lack of Key Managers in the National Office

Initiatives are under way at the national office to strengthen monitoring
of the rental assistance distribution process at all levels. In response
to the Improper Payments Information Act, the national office is
conducting a review of properties to determine a nationwide erroneous
rental assistance payments rate.14 The national office is also attempting
to gain access to the National Directory of New Hires, which is
administered by the Department of Health and Human Services. This database
would give RHS state and local offices access to national wage-matching
data that could be used to further verify tenants' income.

A number of key positions in the Rural Development mission area,
particularly in the Multifamily Housing Division, have been unfilled since
approximately March 2003, and existing staff were filling two or more
administrative roles (fig. 5).

14The Improper Payments Information Act of 2002 requires that agencies
review all their programs and activities annually and identify those that
may be susceptible to significant improper payments. The act further
requires those agencies with improper payments exceeding $10 million to
provide a report on the actions being taken to reduce the payments.

Figure 5: Key Management Positions in the Rural Housing Service, as of July 2004

                      Rural Housing Service Administrator

                          (Position filled July 2004)

                    Multifamily Housing Deputy Administrator

(Vacant since March 2003)

Assistant Deputy Administrator

       (Vacant since August 2003) Community Programs Deputy Administrator

Processing Division Portfolio Management Director Division Director

Position permanently filled

Position currently vacant but being filled by another employee

Position currently vacant

Points to other positions employee held in an acting capacity (e.g.,
Portfolio Management Division Director was also acting Assistant Deputy
Administrator at one point)

Source: GAO.

For example, as of July 2004 the Deputy Administrator of Multifamily
Housing and Assistant Deputy Administrator of Multifamily Housing
positions had been unfilled since March 2003 and August 2003,
respectively.15 Additionally, at times the Deputy Administrator for
Community Programs filled the roles of both acting Rural Housing Service
Administrator and acting Deputy Administrator of Multifamily Housing. 16
Finally, the Director of the Portfolio Management Division, which
administers the rental assistance program, had also at one time filled the
role of acting Assistant Deputy Administrator for Multifamily Housing. The
leadership gap could have a negative effect on program staff's ability to

15The Deputy Administrator of Multifamily Housing position was filled in
August 2004.

16Community Programs, a division of the Rural Housing Service, is part of
USDA's Rural Development mission area. Community Programs administers
programs designed to develop essential community facilities for public use
in rural areas.

develop and administer new policies in a timely fashion. For example, the
national office planned to release administrative notices clarifying
procedures on, among other things, income verification in July 2004.
However, because of staffing constraints due to the consolidation of RHS
regulations, the issue date was extended.

RHS's Internal Controls Do Not Provide Adequate Oversight of Rental
Assistance Funds

RHS's internal controls do not provide reasonable assurance that rental
assistance funds are used efficiently or that tenants' incomes and assets
are adequately verified. First, RHS lacks sufficient guidance to oversee
the transfer process, creating difficulty in moving unused rental
assistance throughout the multifamily housing portfolio. Second, the
triennial supervisory review does not provide reasonable assurance that
errors in tenants' income statements and thus in rental assistance
payments will be prevented or detected in a timely manner. The guidance is
inconsistently interpreted and the reviews are conducted infrequently
using methods, specifically the selection of an appropriate sample size to
review, which may be insufficient. Alternate methods of verification, such
as internal MFIS database checks and wage matching, could help improve
RHS's existing internal controls if properly designed and implemented.

Insufficient Guidance on Transfers Limits Ability to Move Unused Rental
Assistance to Properties with the Neediest Tenants

The majority of survey respondents (52 percent) very frequently or fairly
frequently encountered high vacancy rates as a cause of unused rental
assistance. Additionally, when asked to identify the three greatest
challenges to their offices, survey respondents most frequently listed
monitoring and filling unit vacancies as the greatest challenge facing RHS
state and local offices. The current lack of clear guidance on
transferring unused rental assistance may prevent RHS officials from using
transfers as a tool to fill vacancies within and outside of a state.

We found that RHS's guidance on transfers was not sufficient to meet GAO's
Standards for Internal Control in the Federal Government. GAO's standards
state that for an entity to run and control its operations, it must have
relevant, reliable, and timely communications relating to internal as well
as external events. For communications to be effective, information must
flow down, across, and up the organization. RD Instructions state that
transferred rental assistance must be directed to properties with tenants
or applicants that have the greatest need, based on income level. However,
RD Instructions are silent on whether state and local offices should
transfer unused rental assistance to properties within their portfolios or
redirect it

to properties outside their portfolios-either within or outside the state-
where the need is greater. As a result, the guidance is interpreted
inconsistently, potentially preventing RHS from effectively transferring
rental assistance.

During our site visits, an RHS official in one state said that her office
most often conducted transfers within the property owners' portfolios, but
officials in other states prioritized transfers to benefit the most needy
tenants throughout the state. An official noted that RD Instructions did
not explicitly prioritize such transfers. Further, because RD Instructions
do not state whether rental assistance is to be transferred only within
states or among states, three state offices interpreted the RD
Instructions as allowing the transfer of unused rental assistance to other
states with a greater need. However, 76 percent of survey respondents
reported that their office did not have, or they did not know if it had,
formal procedures for conducting state-to-state transfers.

In written responses to our survey questions, certain respondents noted
that RHS guidance on transferring unused rental assistance was
insufficient. For example, one survey respondent noted that states were
currently responsible for transferring unused rental assistance without a
great deal of guidance from the national office. Another respondent noted
that regulatory and instructional requirements for transferring rental
assistance were "not efficient." Yet another noted that RD Instructions
did not outline procedures for transferring unused rental assistance among
properties or states.

Officials in the national office told us that one of their functions was
to encourage the transfer of rental assistance both within and among
states. One official stated that RD Instructions did offer guidance on
where to transfer rental assistance but could not point to specific
details. The official also stated that the goal of the transfer policy was
to provide rental assistance to the tenant with the greatest need.
However, the current RD Instruction on transfers does not provide the
necessary control to ensure that this goal is met. In addition, to
determine where the greatest need exists, RHS national office would have
to consistently assess need within portfolios, within states, and among
states. RHS officials acknowledged that they were reviewing the guidance
on transfers and stated that RHS planned to issue an administrative notice
to clarify procedures for transferring unused rental assistance.

The issue of unclear guidance on transfers is complicated by the fact that
RD Instructions give property owners certain protections that limit
transfers of unused rental assistance. While RHS can transfer unused
rental assistance on covered units after 12 months, at least one unit must
remain covered at the property. RHS must also notify property owners 30
days in advance of its intent to make a transfer and ensure that the
property owners know their appeal rights. According to RHS officials in
the states we visited, property owners most often instigated transfers
when they occurred, typically to other properties they owned. An official
in one state estimated that property owners initiated 95 percent of
transfers in that area.

We found that RHS officials in one state hesitated to recommend
transferring unused rental assistance outside of the property owner's
portfolio because of the appeals process. The State Director of Rural
Development in that state said that it could take at least 18 months to
transfer unused rental assistance to another property if the property
owner appealed, as the appeals process could take up to 6 months.
Additionally, one survey respondent would have liked to have the authority
to transfer unused rental assistance without giving the property owner
appeal rights. RHS national office officials noted that, since 2001, there
had been only two requests for review of a property owner's appeal. In
both cases, RHS's National Appeals Director decided against the property
owner, and allowed RHS to transfer the unused rental assistance to
properties outside of the property owner's portfolios.

Triennial Supervisory Reviews Are Not Adequate as the Primary Monitoring
Tool for the Rental Assistance Program

Approximately 56 percent of survey respondents reported that they were
able to verify the accuracy of information provided during the tenant
certification process. Of those, 56 percent said they did so during the
triennial supervisory review. Likewise, all of the local offices we
visited during our site visits stated that information provided for tenant
certifications was verified during or in preparation for triennial
supervisory reviews. And approximately 83 percent of survey respondents
reported that triennial supervisory reviews provided them with the
information necessary to effectively monitor properties in their
portfolios to a great or very great extent.

However, we found that the triennial supervisory review was not a
sufficient means of detecting problems such as misreported income,
verification errors, and unauthorized rental assistance payments. GAO's
standards state that internal control should generally be designed to
assure

that ongoing monitoring occurs in the course of normal operations, is
performed continually, and is ingrained in the agency's operations. RHS's
triennial supervisory reviews did not meet this standard because they are
conducted infrequently using a sample size that may be too small to
provide reasonable assurance that all tenant certifications are reviewed
at least once or more than once every several years. RD Instructions
require a sample of at least six tenant household files from a property
with more than six units, and all tenant household files for properties
with six units or fewer. With a sample size this small, however, tenants
could move from a unit they had rented before the local office has a
chance to sample their household file during a supervisory review.

Local offices in the states we visited said that they interpreted RD
Instructions as requiring a 20-or 25-percent sample of tenant household
files. However, we found that staffing and time limitations did not allow
staff to conduct the reviews this way.17 For example, an official in one
local office stated that her office reviewed six tenant household files
for every property, even if 20 percent of the property comprised a greater
number of units. Furthermore, about 20 percent of survey respondents
reported not having sufficient staff to conduct reviews, and about 27
percent reported not having sufficient time to conduct the reviews
according to the RD Instructions. A local office official in another state
we visited said that staff reviewed either six tenant household files or
10 percent of the totalwhichever was greater. Even if the local offices
consistently used a 20 percent or 25 percent sample of tenant household
files, however, as figure 6 shows a rental assistance unit could go
without review for 12 to 15 years.18

17A national office official stated that local offices likely interpret
the Instructions this way because Exhibit F-1 in RD Instruction 1930-C
uses 25 percent as an example sample size. Ultimately, according to the
Instructions, the percentage of units a local office should review is
based on that office's "preference for sample size," as long as local
offices select at least six units for properties with six or more units.

18This figure assumes a 20 percent or 25 percent sampling of tenant
household files. The figure also assumes that RHS local staff sample
tenant household files without replacement. That is, the portion of files
selected on an initial supervisory review will not be selected for
subsequent supervisory reviews.

Figure 6: Frequency of Supervisory Reviews of Tenant Household Files

Source: GAO.

Note: This figure assumes a 20 percent or 25 percent sampling of tenant
household files. This figure assumes that RHS local office staff sample
tenant household files without replacement. That is, the portion of files
selected on an initial supervisory review will not be selected for
subsequent supervisory reviews.

RHS's national office does not have a basis for determining whether
sampling a minimum of six tenant household files per property is
sufficient to provide reasonable assurance that errors in tenant and
payment information are adequately detected. In order to design a cost
efficient sampling strategy that can provide reasonable assurance with
respect to tenant income and rental assistance payments, RHS needs
information on the overall error rates. One way to determine an overall
error rate is to conduct a national statistically valid random sample of
tenant household files. Information gathered from this sample would
provide RHS with a better idea of the level of erroneous income tenants
were reporting and enable RHS to estimate the level of improper rental
assistance payments that are being made throughout the program. In
addition, it would provide RHS officials with valuable information on how
to structure future supervisory reviews in a more efficient and effective
manner.

In a June 2003 report assessing controls over the tenant certification
process and accuracy of rental subsidies conducted by the USDA Office of
Inspector General, the audit team found that in Florida, tenant
certifications for an estimated 20.8 percent of the 14,705 units contained
errors that resulted in households receiving an estimated $4.7 million in

                                  Total units
                 25% sample  25% sample  25% sample  25% sample  

                                  Total units
                      20%       20%       20%       20%      20% 
                   sample    sample    sample    sample   sample 

improper rental subsidies.19 This finding emphasizes the importance of
being able to detect errors in tenant information in a timely manner.

In March 2004, RHS began a review of properties to determine an overall
error rate for rental assistance payments to comply with the Improper
Payments Information Act. RHS officials selected a sample of 539 assisted
units in 321 RHS properties, and the national office expects to report
their findings by the end of September 2004.20 Until RHS determines an
overall error rate using a statistically valid sample, RHS will not be
able to determine an appropriate sampling strategy for use with
supervisory reviews to estimate error rates and compliance with internal
controls.

GAO's internal control standards also state that agencies should evaluate
their monitoring activities, and the evaluations should depend on the
assessment of risks and the effectiveness of the monitoring procedures.
RHS officials in the states we visited said that they found errors or
questionable tenant documentation during triennial supervisory reviews.
For example, RHS officials in one state said that they would follow up
with the property owner if they found errors in the data that are being
verified. The official added that property owners might be given a 15-or
30-day period to correct any deficiencies. An RHS official in another
state said that RHS staff sometimes found questionable supporting
documents during triennial supervisory reviews. The official explained
that the staff note the problem in the triennial supervisory review report
and request that the property owner obtain the correct documentation.
These examples suggest that, if properly designed and implemented, the
triennial supervisory review can be valuable in detecting tenant and
payment errors.

19The 95 percent confidence interval for the 20.8 percent estimate is from
11.3 to 30.2 percent. The estimated $4.7 million, reported in "Rural
Development, Rural Housing Service, Rural Rental Housing Program, Tenant
Income Verification, Gainesville, Florida, Audit No. 04004-3At," consisted
of estimated overpayments of $4.4 million and underpayments of $0.3
million. The 95 percent confidence interval for this total amount extends
from $2.4 to $7.0 million.

20We did not assess the validity of RHS's sample selection methods.

Alternate Controls Used by RHS to Verify Tenant and Payment Information
Have Had Limited Effectiveness

RHS staff use alternate controls, such as MFIS database checks and wage
matching, to verify tenant and payment information. According to RHS
officials, the MFIS system has internal checks that help RHS officials
avoid and catch errors. When tenant certifications are submitted
electronically through the Industry Interface, MFIS automatically rejects
those that are problematic. For example, a tenant certification would be
rejected if it showed that the tenant might already reside at another
property-that is, if the tenant's Social Security number already existed
within the system.

Also, MFIS allows RHS officials to reject tenant certifications for
certain reasons and thus flag them for review. For example, as we have
seen tenants must recertify their eligibility for rental assistance if
their income increases by $40 a month ($480 a year) or decreases by $20 a
month ($240 a year). MFIS allows each office to compare incoming tenant
certifications to the prior batch of tenant certifications. RHS staff in
each office can then establish a threshold in MFIS to reject tenant
certifications that show a change in income of more than an established
amount, for example, 20 percent, and hold them for further review.

Although RD Instructions determine income changes in terms of dollar
amounts, the MFIS database check has been implemented to use percentages
to set the threshold for rejecting tenant certifications. However, setting
the threshold in terms of dollar amounts would be consistent with the RD
Instruction that describes how to determine changes in income. For
example, if a tenant reported zero income on an initial certification, any
subsequent increase in income would be more than 20 percent, thereby
triggering a rejection, although the dollar increase could be less than
$40. Likewise, if a tenant's reported income increased from $30 to $40 per
month, for example, the MFIS system would also reject this certification
because the percentage increase is greater than 20 percent. However, a $10
increase in income is not enough to warrant recertification. But if a
tenant's reported income increased from $300 to $345 per month, MFIS would
accept this certification because the percentage increase in income is
less than 20 percent, although the dollar increase is more than $40.

An RHS official in the national office stated that the MFIS database check
could be designed to use dollar amounts instead of percentages to
establish the threshold. In either case, however, the national office does
not currently have an established percentage benchmark in MFIS for all
offices

to use, and RHS local offices are not setting the thresholds
consistently.21 In one local office we visited, RHS staff established
their threshold to reject tenant certifications that change by 20 percent.
However, in another state we visited, local office staff set MFIS to
reject changes in income of 25 percent or more. Setting the threshold
higher than an established benchmark would make this control less
effective at detecting changes in income. Setting a threshold lower than
an established benchmark would make this control more effective but would
result in a higher number of pending certifications that would likely
require more staff resources.

Although the RHS national office has established the MFIS database check
as a control, we identified and were told of errors on tenant
certifications. For example, in one state we identified a tenant
certification that listed two dependents as living in a unit receiving
rental assistance. However, the dependents were not listed in MFIS. RHS
local office officials stated that, in such a situation, the two
dependents would not change the amount of rent owed by the tenant or the
amount of rental assistance due to the property owner. But because rental
assistance payments are based on household size, among other things, the
presence or absence of dependents could influence rental assistance
payments. For example, if a tenant got married an additional person would
be living in the unit and would possibly have an additional source of
income. If a tenant gave birth, the tenant would receive an income
deduction for the newborn, decreasing the amount of income used in
determining the amount of the rental assistance payment.

Further, RHS officials in four of the states we visited stated that,
following RHS guidance, they verified the accuracy of rental assistance
payments by comparing property owners' project worksheets with project
worksheets generated using MFIS. According to an RHS official, any errors
on the property owners' project worksheets should be detected during the
comparison with the MFIS-generated project worksheet. But since the MFIS
project worksheet is generated with the same information from tenant
certifications that property owners used to generate their project
worksheets, RHS is essentially comparing duplicate data from the same
source (see fig. 7).

21A national office official stated that the benchmark was 10 percent when
this database check was designed, but noted that MFIS had been redesigned
to give local offices the flexibility to establish their own thresholds.

Figure 7: Process Used to Verify the Accuracy of Rental Assistance
Payments

Sources: GAO and Art Explosion (images).

According to GAO Standards for Internal Control in the Federal Government,
such a check is not sufficient. Internal control calls for comparisons and
assessments using different sets of data so that relationships can be
analyzed and appropriate actions taken. In order for RHS to have a
meaningful reconciliation, MFIS data would need to be compared with a
different data source to identify differences in reported tenant and
payment information.

An RHS official in the national office believed that this internal control
standard is not applicable to this situation. The official stated that
discrepancies between the MFIS-generated project worksheet and the
property owner's project worksheet often arise because the property owner
may not send complete information. For example, if a property owner did
not submit a form indicating that a tenant had vacated a unit, or if the
owner did not perform a calculation correctly, a discrepancy would

occur that RHS would look into. We agree that this system can catch errors
of inconsistency. But a property owner could consistently be submitting
complete, but inaccurate, information reported by the tenant. In such
cases, RHS would not be able to detect errors or improper rental
assistance payments by comparing the two project worksheets. RHS would
need to consult another source of information, such as data on earned
income that could be used for wage matching, to verify that information
reported by the tenant and submitted by the property owner is accurate.
Wage matching would allow staff to verify tenant's reported income using
wage information collected by the state.

RHS was told by the USDA Office of the Inspector General over 5 years ago
to continue their efforts to ensure that RHS state and local offices
utilize wage matching to detect misreported tenant income.22 In a February
1999 report, the Inspector General's audit team evaluated methods RHS and
property owners used to ensure that tenants' incomes were properly
reported on tenant certifications. The audit team found that 47 percent of
tenant certifications in six states (29,238 out of 62,815) were incorrect
due to inaccurately reported income, income changes, and property managers
that did not properly verify income reported by tenants. Further, the
audit team found that about 93 percent of the income errors involved
tenants that had reported earned income. The Inspector General recommended
that RHS be more aggressive in expanding its wage-matching efforts.

Some RHS state offices are granted access to wage-matching data through
agreements between the offices and wage collection agencies, such as state
labor departments. These agreements follow state laws. Our survey results
showed that 64 percent of RHS state offices (28 of 44 RHS state offices
that responded) had access to state employment information to conduct wage
matching (see fig. 8.)

22USDA Office of the Inspector General Evaluation Report No.
04801-0004-Ch, Rural Housing Service, Rural Rental Housing Program, Tenant
Income Verification Process, Washington, D.C., Feb. 1999.

Figure 8: States with Wage-Matching Agreements

Source: GAO.

Note: In certain instances, RHS has a single state office for selected
groupings of states. To depict wage-matching status in these selected
state groupings, this figure only shades the state where the state office
is located. See appendix I for more details on RHS's state groupings.

However, some RHS officials believe that state-based wage matching has
limitations. For example, officials in one state we visited noted that
their office did not place much emphasis on wage matching because their
area bordered other states and many individuals worked across state lines.
Without access to wage-matching data in the bordering states, RHS staff
were not able to adequately certify reported income. Further, officials
noted that state wage data were often outdated and did not capture all
sources of income.

RHS is attempting to acquire access to the Department of Health and Human
Services' National Directory of New Hires by contacting the U. S.
Department of Housing and Urban Development (HUD) to discuss the

language originally proposed by HUD for the authority to use the
directory.23 The National Directory of New Hires, which includes
centralized sources of state wage, unemployment insurance, and new hires
data for all 50 states, would provide nationwide data for wage matching.
Access to this directory would likely help solve the problem of verifying
income for tenants who work across state lines. Also, wage data from the
National Directory of New Hires is more recent than state data and
captures all earned income.

Conclusions	RHS's internal controls do not provide reasonable assurance
that rental assistance resources are used effectively. RHS national office
transfer guidance is insufficient and RHS local offices have
inconsistently interpreted what is available. As a result, rental
assistance funds may not be directed to the most needy tenants or may go
unused. Additionally, the infrequency of the triennial supervisory reviews
using a sampling methodology that may be insufficient cannot reasonably
ensure that tenants' income and assets, and ultimately rental assistance
payments, are adequately verified. RHS does not have a basis for
determining what constitutes a sufficient sample size for the triennial
supervisory reviews without establishing an overall error rate for rental
assistance payments. These shortcomings increase the risk that RHS will
provide rental assistance to tenants that may not be eligible.

RHS currently uses alternate methods of income verification, including
MFIS database checks that, if properly designed and implemented, could
provide reasonable assurance that errors in tenant and payment information
are detected in a timely manner. However, the effectiveness of these
alternate methods is limited because of their faulty design and
implementation. RHS officials stated that access to the National Directory
of New Hires would help with income verification and has made efforts to
gain access to this information.

23Through Section 453 (j)(7) of the Social Security Act, "Information
comparisons for housing assistance programs," 42 USC 653 (j)(7), HUD was
given the authority to access nationwide tenant wage information through
the National Directory of New Hires. According to RHS, its request for
inclusion, however, came too late.

Recommendations for Executive Action

To ensure that rental assistance funds are effectively distributed to
properties that have tenants with the greatest need, GAO recommends that
the Secretary of Agriculture require program officials to (1) establish
centralized guidance on transferring unused rental assistance, (2) improve
sampling methods to ensure that a sufficient number of tenant household
files are selected for supervisory reviews and (3) improve verification of
tenant information including more effective use of alternate methods of
income verification.

Matter for Congress may wish to consider adding RHS to Section 453 (j)(7)
of the

Social Security Act, "Information comparisons for housing
assistanceCongressional programs."24 This action would grant RHS access to
the Department of Consideration Health and Human Services' National
Directory of New Hires for wage

matching purposes. If such access were granted, RHS and the Department

of Health and Human Services would have to develop specific procedures

to facilitate it.

Agency Comments and Our Evaluation

We provided USDA with a draft of this report for review and comment. The
Acting Undersecretary for Rural Development provided written comments that
are presented in appendix IV. The letter included a summary of activities
that were planned or had recently begun at RHS that the Acting
Undersecretary said were designed to improve the oversight, monitoring,
and guidance of the multifamily housing program.

In response to our recommendation on establishing centralized guidance on
transferring unused rental assistance, the Acting Undersecretary for Rural
Development agreed to establish this guidance in an administrative notice.
Concerning our recommendation that RHS improve sampling methods to ensure
that a sufficient number of tenant files are selected for supervisory
reviews, the Acting Undersecretary responded that USDA would instead
revamp the supervisory review process to improve its effectiveness and
consistency. We agree that the supervisory review process should be
revised to improve its effectiveness and consistency and are encouraged
that USDA is taking steps to accomplish this. However, to make this
process an effective control to ensure that the rental assistance

2442 USC 653 (j)(7).

payments are correct, or that incorrect payments are discovered in the
normal course of business, USDA must also improve its sampling methodology
as we described in the report. In particular, RHS must determine an
overall error rate using a statistically valid sample to determine how
many tenant household files are sufficient to detect errors. Unless this
methodology is improved in conjunction with the supervisory review
process, USDA will continue to lack any assurance as to the propriety of
the rental assistance payments that are not reviewed. The Acting
Undersecretary agreed with our recommendation that RHS improve its methods
for verifying tenant information, including using alternate income
verification methods. He stated that USDA would institute broader use of
alternate income verification methods, such as improved access to wage
matching and new hires data, but noted that USDA needed to be able to
share this data with third parties, particularly the property owners. As
noted in this report, if RHS were granted access to data from the National
Directory of New Hires, RHS and the Department of Health and Human
Services would have to develop specific procedures to facilitate such
access.

The Acting Undersecretary also noted that GAO said in the report that
streamlining and automating borrower's requests for rental assistance
payments could remove an important layer of agency review. However, GAO
did not take a position on this issue. We merely summarized concerns
expressed by RHS local office staff we visited, attributing the comments
appropriately. We also noted that national office staff overseeing the
changes had told us that streamlining the payment process would not
eliminate this review and that not all states were currently using the
streamlined process. Finally, we incorporated updated information USDA
provided on RHS staffing.

As agreed with your office, unless you publicly announce the contents of
this report earlier, we plan no further distribution until 30 days from
the report date. At that time, we will send copies of this report to
interested Members of Congress and congressional committees. We will also
send copies to the Secretary of the Department of Agriculture and the
Director of the Office of Management and Budget and make copies available
to others upon request. In addition, this report will be available at no
charge on the GAO Web site at http://www.gao.gov. Our survey of RHS state
and local offices administering Section 521 rental assistance and a more
complete tabulation of the survey results (GAO-04-978SP) will also be
available on the GAO Web site at www.gao.gov/cgi-bin/getrpt?GAO-04978SP.

Please contact me at (202) 512-4325, or Andy Finkel at (202) 512-6765, if
you or your staff have any questions concerning this report. Key
contributors to this report are listed in appendix V.

Sincerely yours,

William B. Shear Director, Financial Markets and Community Investment

Appendix I

Scope and Methodology

To determine how RHS distributes rental assistance funds and monitors the
rental assistance program, and to identify the internal controls RHS has
established to ensure that rental assistance funds are administered and
used effectively, we collected written and testimonial information from
agency officials in the national office and in the Office of the Deputy
Chief Financial Officer in St. Louis, Missouri on the rental assistance
distribution process and the guidance governing the distribution process.
We consulted GAO's Standards for Internal Control in the Federal
Government (GAO/AIMD-00-21.3.1) to review control activities that apply to
RHS's distribution of rental assistance funds. We used these standards to
assess whether RHS's current internal controls were sufficient to
effectively monitor the rental assistance program.

We also surveyed RHS state and local offices that were identified as
having rental assistance responsibilities by their respective State
Directors or the appropriate designee. We identified 248 state and local
offices with rental assistance responsibilities.1 These consisted of 46
RHS state offices and 202 local offices in existence as of December 2003.2
To ensure security and data integrity, we provided each of the 248 offices
with a web address, a unique identification number, and a password to
allow the office access to one questionnaire.

The questionnaire covered topics related to the rental assistance program.
Specifically, the questions covered general office and staff information,
renewal contracts, unused rental assistance, transfers, tenant
certifications, project worksheets, rent setting, reserve accounts, and
technology and resources. To prepare for the survey, we conducted
interviews in three state offices: the combined state office of Maryland
and Delaware, and the Texas and Illinois state offices, and interviewed
RHS local offices in Texas and Illinois. Of these, Maryland and Delaware
and Texas had centralized their processing of rental assistance payments.
We also pretested the questionnaire with state and local offices in
Pennsylvania and Wisconsin to determine whether respondents would

1Consistent with its use throughout this report, the term "local office"
encompasses RHS local, area, servicing, and field offices.

2The following state groupings have a combined state office: Delaware and
Maryland; Florida and the Virgin Islands; Hawaii, American Samoa and the
Western Pacific; Massachusetts, Connecticut and Rhode Island; and Vermont
and New Hampshire. Also, we only surveyed state offices with rental
assistance responsibilities. Thus, the number of state offices we surveyed
was less than 50.

Appendix I Scope and Methodology

understand questions the way we intended. Since the questionnaire was
administered via the Internet, we conducted a usability pretest with yet
another RHS local office in Virginia to observe respondents answering the
questionnaire as it would appear when activated.

We took the following steps to increase the response rate for all sectors
of the population. We sent one reminder notice via e-mail and conducted
follow-up telephone calls to those offices that did not respond to our
survey by the initial deadline. Collection of survey data ended on May 3,
2004. We received responses from 230 offices, providing an overall
response rate of 92.7 percent. Specifically, 44 state offices and 186
local offices responded, resulting in response rates of 95.7 percent and
92.1 percent, respectively. We did not independently verify the
information obtained through the survey.

The practical difficulties of conducting any survey may introduce certain
types of errors, commonly referred to as nonsampling errors. For example,
differences in how a particular question is interpreted, the sources of
information available to respondents, or the types of people who do not
respond can introduce unwanted variability into the survey results. Steps
such as pretesting and follow-up contacts to increase response rates serve
to minimize nonresponse error. In addition, steps such as performing
statistical analyses to identify inconsistencies and having a second
independent reviewer for the data analysis can further minimize such
error.

Data from the Web survey were entered electronically by participants and
imported into an electronic data file. Close-ended questionnaire items
were analyzed using statistical software. A coding scheme was developed,
and open-ended questionnaire items were content analyzed by two trained
coders, who revised their codes until 100 percent agreement was reached.
The results of the content analysis apply only to those survey
participants who provided responses. We only content analyzed open-ended
responses that were discussed in this report. We conducted our survey work
from April 2003 through July 2004 in accordance with generally accepted
government auditing standards.

To obtain more in-depth information, we conducted site visits to RHS state
and local offices in five states: Utah; South Carolina; New York;
Mississippi; and the Connecticut, Massachusetts, and Rhode Island state
grouping. We designed a site visit interview guide that enabled the team
to follow the distribution process beginning with the property owner or
manager and proceeding to the local office, and finally, to the state
office. Using the

Appendix I Scope and Methodology

interview guide, we observed and discussed with RHS staff the tenant
certification, project worksheet, and transfer processes, among others.

To determine the states where we would conduct the site visits, we
conducted a stratified random sample. Two centralized and three
decentralized states were randomly selected from a list of state offices,
after eliminating the offices we visited to prepare and pretest the
survey.3 After the state offices were selected, we randomly selected one
local office from the local offices available in that state. We
judgmentally selected the property owners based on their distance from the
local office and their most recent rental assistance transfer. However,
the experiences of these states are not necessarily representative of the
experiences of any other state.

During each of the site visits, we interviewed a property owner or
manager, RHS officials in a local office, and RHS officials in the state
office. Using the interview guide, we also obtained key documents related
to the three processes that were being discussed. Specifically, the
interview guide required the team to obtain from the property owner or
manager the most recent, completed project worksheet; tenant
certifications associated with certain rental assistance units; tenant
applications corresponding to the tenant certifications; a current waiting
list; and documentation of a recent transfer of rental assistance. We
tracked these documents through the local offices and the state offices to
observe internal controls at each level. After all site visits were
completed, we interviewed officials from RHS's Office of the Deputy Chief
Financial Officer to obtain information on their role in the rental
assistance distribution process. Last, we interviewed the RHS national
office to determine its role in monitoring the rental assistance
distribution process and to assess the internal controls they have
established to oversee rental assistance funds.

We also consulted GAO's Standards for Internal Control in the Federal
Government to review control activities that apply to RHS's distribution
and monitoring of rental assistance funds. We reviewed site visit
documents to provide examples of RHS's internal controls. We performed our
site visit work from March through May 2004 in accordance with generally
accepted government auditing standards.

3During our site visits, we learned that one of the two centralized states
selected was, in fact, a decentralized state for rental assistance payment
processing.

Appendix II

State Office Survey Responses for Rural Rental Housing (RRH) Properties

                                                     RRH units 
                                     RRH units occupied by     
                                               rent-           
                                   occupied by    overburdened 
                                          zero                 
                                 income         tenants, as of      RRH units 
                                 tenants, as                        receiving 
State                     RRH                                       rental 
         RRH properties    units     of 1/1/04          1/1/04     assistance 

          Alaska           42         897        32        44     
         Alabama           495      16,191      954      3,206          8,033 
         Arkansas          361       9,908      174      1,076          6,092 
         Arizona           118       3,775      140       217           3,080 
         Delaware          221       6,975      581       320           4,210 
         Georgia           478      15,427      233      4,134          7,972 
          Hawaii           22         903        3         3      
           Iowa            679      11,717      242       170           8,884 
          Kansas           409       6,651      111       613           3,858 
         Kentucky          456      12,284      291       525           6,437 
        Louisiana          415      12,637      347      2,360          7,537 
      Massachusetts        143       4,949      491       349           3,487 
         Michigan          723      18,653       NR        NR           8,905 
        Minnesota          673      12,064       DC        DC           6,389 
         Missouri          896      19,868       DK      1,964          8,493 
       Mississippi         564      16,416      669      1,252          8,825 
         Montana           168       2,634       87       184           1,939 
       North Dakota        253       3,365      187       102           2,147 
         Nebraska          280       3,908      197       141           2,566 
        New Jersey         83        3,347       78       445           2,009 
        New Mexico         116       4,005      347       211           3,072 
          Nevada           80        1,800       DK        DK           1,100 
         New York          471      13,210      148      1,101          5,194 
           Ohio            402      14,912      553      1,429          8,905 
         Oklahoma          302       8,111      224       140           5,130 
       Pennsylvania        330      10,640       DC        DC           6,840 
      South Carolina       352      12,309      430      2,847          5,880 
       South Dakota        503       6,844       DK        70           4,570 
        Tennessee          386      13,038       DK        DK           6,539 
          Texas            804      26,466      707      4,046         14,397 
           Utah            93        2,095       DK        DK           1,595 
         Virginia          272      10,498       DC        DC           6,461 

                                  Appendix II
                    State Office Survey Responses for Rural
                        Rental Housing (RRH) Properties

                         (Continued From Previous Page)

                       Washington 298 8,989 97 909 6,205
      RRH       RRH              occupied                  income   tenants, RRH units                    RRH                   rental 
    units     units occupied      by zero overburdened     tenants,    as of receiving State        RRH units     of        assistance 
                    by rent-                               as                                properties       1/1/04 1/1/04 

                     West Virginia 255 7,221 281 977 4,240

                         Wyoming 62 1,655 118 252 1,188

Source: GAO.

Notes:

DC = Don't collect

DK = Don't know

NR = No response

We surveyed only state offices with rental assistance responsibilities.
Also, in certain instances, RHS has a single state office for selected
groupings of states. For these selected groupings, this chart only
displays the state name where the state office is located. Thus, the total
number of state offices is less than 50. See appendix I for more details
on RHS's selected state groupings.

This chart displays survey responses for RHS state office respondents
only, if they indicated that they had Rural Rental Housing (Section 515)
properties in its portfolio and if they provided data other than Don't
Collect, Don't Know, or No Response for at least one of the above
questions.

Appendix III

State Office Survey Responses for Off Farm Labor Housing (FLH) Properties

Off FLH units

                     occupied by zero Off FLH units occupied by Off FLH units
               income tenants, as rent-overburdened tenants, receiving rental
     State Off FLH properties Off FLH units of 1/1/04 as of 1/1/04 assistance

Hawaii 131 0 0

Massachusetts 2 48 2 1

        Michigan              2      44          0             0 
       New Mexico             5     111         19             5 
        Oklahoma              2      42          0             0 
         Texas               20    1,415        13            36        1,000 
        Virginia              2      46         DK        DK     
       Washington            24     753          9            21 

Source: GAO.

Notes:

DK = Don't know

We surveyed only state offices with rental assistance responsibilities.
Also, in certain instances, RHS has a single state office for selected
groupings of states. For these selected groupings, this chart only
displays the state name where the state office is located. Thus, the total
number of state offices is less than 50. See appendix I for more details
on RHS's selected state groupings.

This chart displays survey responses for RHS state office respondents
only, if they indicated that they had Off Farm Labor Housing (Section
514/516) properties in its portfolio and if they provided data other than
Don't Collect, Don't Know, or No Response for at least one of the above
questions.

Appendix IV

Comments from the U.S. Department of Agriculture

Appendix IV
Comments from the U.S. Department of
Agriculture

Appendix IV
Comments from the U.S. Department of
Agriculture

Appendix V

                     GAO Contacts and Staff Acknowledgments

GAO Contacts	Andy Finkel (202) 512-6765 [email protected] Katherine Trimble
(202) 512-5033 [email protected]

Staff 	In addition to those named above, William Bates, Emily Chalmers,
Jamila Jones, Austin Kelly, Alison Martin, Marc Molino, and Julie Trinder
made key

Acknowledgments contributions to this report.

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

Obtaining Copies of The fastest and easiest way to obtain copies of GAO
documents at no cost

is through GAO's Web site (www.gao.gov). Each weekday, GAO postsGAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To Testimony have GAO e-mail you a list of newly posted products
every afternoon, go to

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

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

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

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

To Report Fraud, Contact:
Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm

E-mail: [email protected] Federal Programs Automated answering system:
(800) 424-5454 or (202) 512-7470

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

Relations Washington, D.C. 20548

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

                               Presorted Standard
                              Postage & Fees Paid
                                      GAO
                                Permit No. GI00

United States Government Accountability Office Washington, D.C. 20548-0001

Official Business Penalty for Private Use $300

Address Service Requested
*** End of document. ***