This is the accessible text file for GAO report number GAO-04-424 
entitled 'Rural Housing Service: Standardization of Budget Estimation 
Processes Needed for Rental Assistance Program' which was released on 
March 25, 2004.

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

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

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

March 2004: 

RURAL HOUSING SERVICE: 

Standardization of Budget Estimation Processes Needed for Rental 
Assistance Program: 

GAO-04-424: 

GAO Highlights: 

Highlights of GAO-04-424, a report to the Ranking Minority Member, 
Subcommittee on Agriculture, Rural Development, and Related Agencies, 
Senate Committee on Appropriations 

Why GAO Did This Study: 

The Rural Housing Service’s (RHS) Section 521 Rental Assistance 
Program provides rental subsidies to about 250,000 rural tenants. With 
an annual budget of over $700 million, the program is RHS’s largest 
line-item appropriation, accounting for approximately 70 percent of 
the agency’s budget. In early 2003, RHS reported hundreds of millions 
of dollars in unexpended balances, primarily tied to 5- and 20-year 
contracts issued from 1978 through 1982. Concern has arisen that these 
unexpended balances may be the result of the agency’s budget 
practices, especially its procedures for estimating funding needs. GAO 
was asked to assess the accuracy of RHS’s budget estimates for the 
rental assistance program, the activity level of rental assistance 
contracts issued from 1978 through 1997, and the activity level of 
rental assistance contracts issued from 1998 through 2002 and the 
accuracy of RHS’s estimates of the rate at which these funds would be 
used.

What GAO Found: 

RHS is overestimating its budget needs for 5-year rental assistance 
contracts in three ways. First, the agency uses inflation factors that 
are higher than those OMB recommends for use in the budget process. 
Second, RHS does not apply its inflation rate separately to each year 
of a 5-year contract, but instead compounds the rate to reflect the 
price level in the fifth year and applies that rate to each contract 
year. Using these first two methods, RHS overestimated its 2003 budget 
needs by $51 million (6.5 percent). Third, RHS bases its estimates of 
future expenditure rates on recent maximum expenditures, rather than 
on the average rates at which rental assistance funds are expended. 
RHS has begun the process of automating its budget processes and 
certain aspects of its new model promise improvements over the current 
estimating methods. However, the agency continues to use its own 
inflation rates and incorrectly calculates those rates in such a way 
that would cause the agency to actually underestimate its budget 
needs.

At current spending rates, it will take another 7 years for all the 
active contracts that were issued from 1978 through 1982 to expend 
their funds, 8 years after the last of the 20-year contracts were 
expected to expire. Contracts issued from 1983 through 1997 should 
expend their remaining funds in 2004. GAO calculated that RHS 
overestimated its funding needs for contracts issued from 1998 through 
2002 by an average of about 8 percent each year. GAO analysis of 
rental assistance payment data showed that the agency has 
overestimated its budget needs almost every year since 1990, the 
earliest year for which GAO gathered data. Where GAO had sufficient 
data from the agency, the analysis also shows that if RHS had used and 
correctly applied OMB inflation rates to its base per-unit rates, its 
estimates would have been closer to actual expenditures. 

Standardizing the agency’s budget estimation processes would help the 
agency more accurately estimate its rental assistance needs and 
curtail future unexpended balances or budget shortfalls.

What GAO Recommends: 

To more accurately estimate rental assistance budget needs, GAO 
recommends that the Secretary of Agriculture require program officials 
to use and correctly apply the inflation rates provided by the Office 
of Management and Budget (OMB) in its annual budget estimation 
processes. The Department of Agriculture and OMB commented on a draft 
of this report.

www.gao.gov/cgi-bin/getrpt?GAO-04-424.

For more information, contact William B. Shear at (202) 512-4325 or 
shearw@gao.gov.

[End of section]

Contents: 

Letter: 

Results in Brief: 

Background: 

RHS's Current Methods Have Overestimated the Agency's Budget Needs by 
as Much as 7 Percent, or $51 Million Per Year: 

Unexpended Funds Primarily from Contracts Issued from 1978 through 1982 
Will Not Be Exhausted until 2011: 

The Activity of Contracts Issued in 1998 through 2002 Is Consistent 
with Earlier Years: 

Conclusions: 

Recommendation for Executive Action: 

Agency Comments and Our Evaluation: 

Scope and Methodology: 

Appendixes: 

Appendix I: Comments from the U.S. Department of Agriculture: 

GAO Comments: 

Appendix II: GAO Contacts and Staff Acknowledgments: 

GAO Contacts: 

Staff Acknowledgments: 

Table: 

Table 1: Differences between RHS and OMB Inflation Rates, 2000-2003: 

Figures: 

Figure 1: Example of RHS Budget and Allocation Estimation Processes: 

Figure 2: Comparison of RHS and OMB-Based Budget Estimation Methods: 

Figure 3: Active v. Expired Contracts and Expended v. Unexpended Amounts 
from Contracts Issued from 1978 through 1982: 

Figure 4: Projected Expenditures from Contracts Issued from 1978 through 
1997: 

Figure 5: Active v. Expired Contracts and Expended v. Unexpended Amounts 
from Contracts Issued from 1998 through 2002: 

Figure 6: Actual v. Estimated Rental Assistance Expenditures Per-Unit, 
Per-Year, from 1990 through 2003: 

Abbreviations: 

AMAS: Automated Multi-Housing Accounting System: 

AMI: Area Median Income: 

OMB: Office of Management and Budget: 

RHS: Rural Housing Service: 

USDA: United States Department of Agriculture: 

Letter March 25, 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: 

Each year, the Rural Housing Service (RHS) of the Department of 
Agriculture (USDA) provides rental subsidies through the Section 521 
Rental Assistance Program to about 250,000 rural tenants living in 
federally subsidized properties. With an annual budget of over $700 
million, the program is RHS's largest line-item appropriation, 
accounting for approximately 70 percent of the agency's budget. In 
early 2003, RHS reported hundreds of millions of dollars in unexpended 
balances, primarily tied to 5-and 20-year contracts issued from 1978 
through 1982. Concern has arisen that these unexpended balances may be 
the result of the agency's budget practices, especially its procedures 
for estimating funding needs.

To help with your oversight of the Section 521 Rental Assistance 
Program, you asked us to assess (1) the accuracy of RHS's budget 
estimates for the rental assistance program, (2) the activity level of 
rental assistance contracts issued from 1978 through 1997 that have 
unexpended balances and the possibility of deobligating[Footnote 1] 
these unexpended balances, and (3) the activity level of rental 
assistance contracts issued from 1998 through 2002 and the accuracy of 
RHS's estimates of the rate at which these funds would be used.

To respond to these objectives, we collected written information from 
RHS, interviewed their representatives, and received oral verification 
from Office of Management and Budget (OMB) staff on RHS's budget 
estimation processes. We reviewed federal guidance on program budget 
preparation and on internal controls that apply to the process. Also, 
we collected and analyzed raw data from RHS's accounting database to 
determine rental assistance program activity. Details about our scope 
and methodology appear at the end of this letter.

We conducted our work from January 2003 through February 2004 in 
Washington, D.C. and St. Louis, Missouri, in accordance with generally 
accepted government auditing standards.

Results in Brief: 

Using its current processes, we found that RHS has overestimated its 
budget needs for 5-year rental assistance contracts in three ways. 
First, the agency has used inflation factors that are higher than those 
OMB has recommended for use in the budget process. Second, RHS does not 
apply its inflation rate separately to each year of a 5-year contract, 
but instead compounds the rate to reflect the price level in the fifth 
year and applies that rate to each contract year. In 2003, for example, 
RHS started with an inflation rate that was high relative to OMB's 
rates and then compounded it, resulting in a budget overestimation of 
over $51 million, or 6.5 percent. Third, RHS bases its estimates of 
future expenditure rates on recent maximum expenditures, rather than on 
the average rates at which rental assistance funds are expended. We 
also found that the budget estimation and allocation processes are 
largely managed by a single employee, which suggests a problem with 
RHS's internal controls for the program. RHS is developing an automated 
budget forecasting and allocation model that shows more promise than 
the current methods it will replace, because it will allow the agency 
to use more accurate data to establish per-unit costs. RHS has also 
taken steps to address the weaknesses in its internal controls and has 
reported that three or four employees, rather than one, will administer 
the new budget forecasting and allocation processes. However, the 
agency plans to continue using its own inflation rates and, in contrast 
to its current methods, will calculate those rates in such a way that 
would cause the agency to actually underestimate its budget needs.

About 18 percent of the rental assistance contracts issued from 1978 
through 1997 are still active--even though the agency estimated that 
they would expire by 2002--and together they account for $605 million 
in unexpended balances. Almost 84 percent of this amount, or $510 
million, involves the 5-and 20-year contracts issued from 1978 through 
1982. Nearly one-third of those contracts are still active. At current 
average spending rates, it will take another 7 years for all the active 
contracts to expend their funds--8 years after the last of the 20-year 
contracts wereexpected to expire. USDA stated that the unexpended 
balances cannot be deobligated, since every rental assistance contract 
RHS has executed contains a provision stating that the agreement 
expires only when the assistance runs out.[Footnote 2]

Ninety percent of the 5-year contracts issued from 1998 through 2002 
are still active, and RHS's estimates of the contracts' expenditure 
rates were inaccurate to varying degrees. Seventy-four percent of the 
total number of contracts issued in 1998 are still active, even though 
RHS expected these contracts to run out during 2003, suggesting that 
these contracts may have been overfunded. About 25 percent, or $114 
million, of the funds remain from the 1998 contracts, and about 35 
percent, or $208 million, remain from the 1999 contracts. Furthermore, 
only 11 percent of the funds from the contracts issued in 2002 were 
spent during the contracts' first 1½ years, suggesting that many of the 
contracts from this time period are expending their funds more slowly 
than RHS anticipated. Our analysis of RHS data found that these 
contracts will run an average of over 6 rather than 5 years each.

To more accurately estimate rental assistance needs and curtail future 
unexpended balances--or budget shortfalls--we recommend that the 
Secretary of Agriculture require program officials to use and correctly 
apply the inflation rates provided by OMB in their annual budget 
estimation and allocation processes.

Background: 

The Section 521 Rental Assistance Program, started in 1978, is 
administered by RHS's Multifamily Housing Division of Portfolio 
Management. 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.[Footnote 3] 
Under the program, eligible tenants pay no more than 30 percent of 
their income toward the rent, and RHS pays the balance to the project 
owner.[Footnote 4] As of January 2003, approximately 53 percent of 
tenants in the program's 464,604 housing units were receiving rental 
assistance. According to program officials, the program has a waiting 
list of approximately 80,000 eligible tenants.

RHS provides the subsidies through 5-year contracts with project 
owners; 20-year contracts were also issued to units in newly 
constructed properties from 1978 through 1982. The contracts specify 
that owners will receive payments on behalf of tenants in a designated 
number of units at the project. Contracts may be renewed as many times 
as funds are made available, and additional units may be covered if 
funds are available. According to program officials, about 96 percent 
of the Rental Assistance Program's budget is used to renew expiring 
rental assistance contracts. The remaining funds are used to provide 
rental assistance for units in newly constructed properties and 
additional rental assistance at existing properties. Budget needs are 
estimated assuming a 5-year rental assistance contract life, although a 
contract's actual life is determined by how long its funds last and 
could run far beyond its estimated life if the funds are expended 
slowly enough.

Each month, project owners or their management companies must certify 
the number of rental assistance units that are occupied. If a unit is 
empty and rental assistance is not being used, the project owner 
assigns a new tenant from the waiting list. If there are no tenants 
eligible for the rental assistance, the rental assistance may be 
transferred to another property.

RHS's national, state, and local offices manage the rental assistance 
program. The national Office of Multifamily Housing Portfolio 
Management develops and implements the program regulations, 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 responsibilities include 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 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. The 
Information Resources Management Office's Rural Housing Service Branch 
maintains the automated databases 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.

In 1982--the fourth year of the program--RHS reported in an internal 
position paper that rental assistance funds were being substantially 
underused. The agency found that approximately $100 million of the 
rental assistance funds obligated for 5-year contracts would be lost 
between 1983 and 1985, because the contracts would expire before all 
the obligated funds were used.[Footnote 5] The study found that rental 
assistance contracts set to expire at the end of 5 years still had 
funds available for an average of 5 additional years. The agency noted 
that if contract terms were extended until the funds ran out, the 
tenants could receive benefits twice as long without any further 
appropriation of funds. Alternately, the agency noted that terminating 
contracts at the end of their terms and returning the unexpended funds 
would save federal funds, assuming the contracts would not be replaced. 
The paper recommended the indefinite extension of rental assistance 
contracts, and the recommendation was adopted as agency policy in 1983. 
Contract language was changed at that time, and previously written 
contracts were amended.

RHS's Current Methods Have Overestimated the Agency's Budget Needs by 
as Much as 7 Percent, or $51 Million Per Year: 

Using its current methodology, RHS has overestimated its budget needs 
for 5-year rental assistance contracts in three ways. First, the agency 
has used inflation factors that are higher than those projected by OMB 
for use in the budget process. Second, RHS compounds the inflation rate 
to reflect the price level in the fifth year and applies that rate to 
each contract year, rather than using an appropriate rate for each 
year. Third, the expenditure rates RHS uses to estimate budget needs 
may also overstate the need for rental assistance. Furthermore, RHS 
budget processes do not adhere to certain internal control standards. 
While a new budget forecasting model shows promise, it is, at present, 
flawed.

Current Processes for Estimating Budget and Allocation Needs: 

RHS's processes for estimating its budget and allocation needs have 
evolved over time. An agency official who worked on the program's 
initial budgets from 1978 through 1982 told us that RHS intentionally 
overfunded the contracts in an effort to subsidize the poorest possible 
tenant by basing tenant contributions on minimum Social Security 
payments. Agency documents suggest that the agency was using inflation 
rates of 10 to 20 percent to estimate future spending rates for the 
life of the contracts.[Footnote 6] After this time period, the agency 
made a series of changes to its processes, including increasing tenant 
contributions from 25 percent to 30 percent, and differentiating per-
unit costs for rural rental housing and farm labor housing properties.

The current method for estimating budget needs was developed by two 
agency officials, between 1995 and 1997, and is based on a formula that 
consists of multiplying estimates for the number of expiring rental 
assistance units by a national average per-unit cost and by an 
inflation factor. The need for rental assistance for units in newly 
constructed properties and additional rental assistance at existing 
properties is also calculated. Since 1996, one official has largely 
managed the process with oversight from the Rural Development Budget 
Office, OMB staff, and RHS supervisors.

OMB Circular A-11, Preparation, Submission, and Execution of the 
Budget, provides guidance to agency officials, stating that preparation 
of agency budgets must be consistent with the economic assumptions 
provided by OMB. These assumptions are listed each year in the 
President's Budget, though they are made available to agencies prior to 
that time for their budget preparations. However, neither OMB staff nor 
the Rural Development Budget Office official we spoke with required RHS 
to use these rates or objected to the agency determining its own 
inflation rates. Although OMB policy does permit agencies to consider 
other factors in developing their budget estimates, it does not allow 
agencies to automatically apply these factors to their budget requests. 
Also, although OMB staff and a Rural Development budget official review 
and approve RHS's budget requests each year, neither has ever required 
nor sought a justification for the rates chosen.

RHS considers its annual rental assistance needs through two different 
processes prior to funding the contracts. First, the agency prepares 
the submission for the President's Budget about 2 years prior to the 
budget year, based on previous renewal needs, average per-unit costs, 
and a compounded inflation rate. Agency officials explained that in 
recent years they used 2.7 percent when preparing estimates to submit 
with the President's Budget. By the time Congress appropriates a budget 
based on this information, the data used by the agency are about 2 
years old. Therefore, after receiving the fiscal year budget, and 
before allocating the rental assistance funds, RHS rechecks the 
estimated number of expiring units and average per-unit costs, based on 
more current data.

For this second process, the agency prepares a report using the 
Automated Multi-Housing Accounting System database (AMAS) on the 
contracts expected to expire in the coming year. This report is sent to 
the local offices for verification and then returned to the national 
office. The agency establishes average per-unit values, based on the 
contracts expiring in the coming year, by inflating the maximum per-
unit cost from the prior 3 months by a compounded inflation rate. 
According to agency officials, RHS adjusts the inflation rate to more 
closely approximate the level of funding received. If the original 
budget estimate and the allocation estimate differ, RHS distributes any 
rental assistance funds remaining after the expiring contracts are 
renewed among states to create new rental assistance units.[Footnote 7] 
These allocation figures then become the basis for future submissions 
for the President's Budget.

Figure 1: Example of RHS Budget and Allocation Estimation Processes: 

[See PDF for image]

[A] Funds for contracts expiring January-March may be allocated in last 
quarter of previous year.

[B] Maximum per-unit cost from prior 3 months based on contracts 
expiring in coming year.

[End of figure]

RHS's Current Methods Have Overestimated Its Budget Needs: 

Because it has not followed OMB procedures, RHS has overestimated its 
future budget needs. For example, although OMB's fiscal year 2003 
published rates varied between 2.2 and 2.4 percent for 2003 through 
2007, RHS used a rate of inflation of 2.7 percent when preparing its 
budget for 5-year contracts that would be renewed in 2003. Applied to 
the average per-unit, per-year base rate of $3,264, the 2.7 percent 
rate created a difference of $203 per unit over the 5-year contract 
period. Since RHS planned to issue 5-year contracts for 44,652 units 
that year, it overestimated its budget needs by more than $9 million. 
RHS did not keep documentation of the inflation rates it used prior to 
2000 but, as table 1 shows, the agency's inflation rates have been 
higher than OMB's rates for every year for which it has documentation.

Table 1: Differences between RHS and OMB Inflation Rates, 2000-2003: 

Budget year: 2000; 
RHS: for President's budget: 2.7%; 
RHS: prior to allocations: 2.5%; 
OMB: 2.3%.

Budget year: 2001; 
RHS: for President's budget: 2.7%; 
RHS: prior to allocations: 3.0[A] %; 
OMB: 2.5%.

Budget year: 2002; 
RHS: for President's budget: 2.7%; 
RHS: prior to allocations: 3.0[B] %; 
OMB: 2.6%.

Budget year: 2003; 
RHS: for President's budget: 2.7%; 
RHS: prior to allocations: 2.4%; 
OMB: 2.2%. 

Source: GAO analysis of RHS and OMB data.

[A] The 2001 budget documents RHS provided show that RHS used different 
inflation rates when reestimating its budget needs for family (3 
percent compounded to the fourth year) and elderly replacement units (3 
percent compounded to the fifth year) and new construction units (2.7 
percent compounded to the fourth year). When asked, RHS reported that 
this is not a standard practice.

[B] Agency officials told us, and agency documents state, that RHS used 
a rate of 2.7 percent for its 2002 allocation estimates. However, upon 
review of the data provided, we determined that they used a rate of 3 
percent and RHS concurred.

[End of table]

However, it is not so much the inflation rates RHS uses as how it uses 
them that has caused the agency to significantly overestimate its 
budget needs. As we have seen, OMB's economic assumptions provide 
inflation rates for each year, and agencies are expected to use these 
individual rates for each year they are projecting, compounding the 
rates separately for each subsequent year based on the previous year's 
rate. But RHS uses one inflation rate for all 5 years of the contract, 
compounds that rate to the fifth year, and then applies the compounded 
rate to each year of the contract. This practice results in the agency 
using a rate that is more than five times the rate it started with for 
the first year.

For example, rather than applying its 2003 inflation rate of 2.7 
percent to each year, RHS compounded this rate to the fifth year and 
applied the resulting value (14.2 percent) back to each year of the 
contract (fig. 2). Spending was thus assumed to be 14.2 percent higher 
for each of the subsequent 5 years, although its stated inflation 
projection was that prices would be only 2.7 percent higher in the next 
year, about 5.4 percent higher in the following year, and so on. Thus 
RHS multiplied the annual per-unit base rate of $3,264 by over 14 
percent, rather than the 2.7 percent it started with, or the 2.2 to 2.4 
percent projected by OMB. Compared with OMB's rates and procedures for 
calculating future spending, RHS's method created a difference of 
$1,147 per unit. Again, as RHS estimated it would fund contracts for 
44,652 units in 2003, it actually overestimated its budget needs by 
over $51 million (6.5 percent).

Figure 2: Comparison of RHS and OMB-Based Budget Estimation Methods: 

[See PDF for image]

[End of figure]

The expenditure rates RHS uses to estimate budget needs may also 
overstate the need for rental assistance. RHS officials claim that the 
expenditure rates they use--equal to the maximum amount of monthly 
rental assistance used over the previous 3 months--account for vacant 
units at the properties. However, using this method treats vacant units 
as if they were occupied and ignores the impact that a property's 
vacancy rate has on rental assistance usage. For example, a property 
with 10 units (receiving $200 of rental assistance for each unit) may 
have no vacant units during the first month, one vacant unit during the 
second month, and 2 vacant units during the third month. During an 
average month, one of the property's units will be vacant, and the 
property will require $1,800 in rental assistance ($200 from each of 
the nine occupied units). However, using the maximum of 3 months, it 
would appear the property has no vacant units and that rental 
assistance needs are $2,000 a month. The estimation rate would be 10 
percent higher than it should be, because the impact of the vacant 
units on rental assistance usage was ignored. We believe using a 3-or 
12-month average could produce a more accurate picture of usage. We 
discussed this practice with RHS officials, and they replied that they 
used the maximum monthly rate because they did not believe averages 
were accurate for their purposes.

RHS Is Not Adhering to Internal Control Standards: 

Finally, RHS is not adhering to internal control standards regarding 
segregation of duties. A single employee within the agency is largely 
responsible for both the budget estimation and allocation processes for 
the rental assistance program. Furthermore, this employee's work is not 
afforded a deliberative review by the Rural Development Budget Office, 
OMB staff, or RHS supervisors. While the office has assigned staff to 
support this employee, the employee told us that due to staff turnover, 
there was no one available to help with the budget estimation and 
allocation processes.

Internal control is a major part of managing an organization. Our 
Standards for Internal Control in the Federal Government 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.[Footnote 8]

Internal control activities are the policies, procedures, techniques, 
and mechanisms that enforce management's directives and help ensure 
that actions are taken to address risks. Control activities are an 
integral part of an entity's planning, implementing, reviewing, and 
accountability for stewardship of government resources. One very basic 
but essential example of a control activity is the segregation of 
duties. According to the standards, key duties and responsibilities 
need to be divided or segregated among different people to reduce the 
risk of error or fraud. This should include separating the 
responsibilities for authorizing transactions, processing and 
recording them, reviewing the transactions, and handling any related 
assets. No one individual should control all key aspects of a 
transaction or event. Based on our analysis of RHS's budget estimation 
and allocation processes, and our Standards for Internal Control, it is 
our view that the autonomy of this employee is not consistent with 
internal control standards.

Automated Process Is an Improvement, but Problems Remain: 

In March 2003, RHS began the process of automating its budget 
estimation and allocation processes by developing a forecasting model 
that it will use to estimate future budget needs, starting in 2006. A 
team consisting of staff from the national office, state offices, and 
the Information Resources Management Office's Rural Housing Service 
Branch created the model. RHS also used contractors and consulted with 
numerous internal experts. The model was designed to automatically 
calculate rental assistance by projecting renewal needs on a property-
by-property basis and calculating future rental assistance usage 
estimates using a combination of factors, including prior actual usage, 
inflation, the potential for rate increases, and rental assistance 
volatility. These last two factors were dropped from the model when the 
agency determined that their impact on future spending was negligible. 
The forecasting model is currently in the testing phase.

RHS demonstrated its new forecasting model to us in late 2003. Certain 
aspects of the model promise improvements over the current estimating 
methods. For example, the model (1) allows RHS to use property-level 
data rather than the national averages that are currently used to 
establish per-unit rates and (2) determines each property's per-unit 
rate based on the average usage over the prior 12 months, rather than 
the maximum usage of the previous 3 months that is currently used. 
These improvements should allow for more accurate replacement estimates 
based on actual rental assistance usage at each property. Furthermore, 
the model properly applies the inflation rate to each of the 5 years 
for which the agency is forecasting. Program officials also told us 
that three to four staff members will be trained to use the model and 
develop the budget and allocation estimates for the agency, which 
should mitigate the segregation of duties concern.

However, the inflation calculation in the model is flawed. RHS 
continues using its own inflation rates rather than those provided by 
OMB, and the agency is incorrectly calculating the rates it plans to 
use. RHS officials explained that they are using historical rates of 
change to determine future spending rates rather than OMB's rates, 
which are based on future projections of inflation. This means that RHS 
is looking to past activity to determine what will happen in the 
future, whereas OMB asks agencies to use projections of future change.

Furthermore, RHS is incorrectly calculating its historical rates of 
change in a way that could cause the agency to underestimate its budget 
needs. RHS calculates the average per-unit expenditures for each year 
over 3 years, then calculates the change from the first year to the 
third and divides that number by 3. The agency should divide by 2, 
since it is estimating an annual rate of change from 2 years of 
changes. By dividing by a larger number than appropriate, RHS's method 
will cause it to underestimate the rate of change and thus to 
underestimate its budget needs. For example, if prices increased 3 
percent from 2000 to 2001, and 2 percent--of the year 2000 level--from 
2001 to 2002, then the average annual increase is 3 plus 2, divided by 
2, which equals 2.5. If RHS divides by 3, it will come up with an 
inflation rate of 1.67--lower than the average inflation rate of the 
recent past.

Unexpended Funds Primarily from Contracts Issued from 1978 through 1982 
Will Not Be Exhausted until 2011: 

Contracts issued from 1978 through 1982 account for the majority of 
unexpended balances and are expending their funds at a relatively slow 
rate. Based on current average expenditures, these contracts likely 
will not expend their funds completely until 2011. USDA has concluded 
that these funds cannot be deobligated. Contracts issued from 1983 
through 1997 also have unexpended balances; based on our analysis, 
these funds likely will be expended in 2004.

Based on their age, contracts issued from 1978 through 1997 (both 5 and 
20 year), should have expired by the end of 2002. As of June 2003, 
approximately 18 percent of these contracts were still active, 
accounting for $605 million in unexpended balances. Most of this amount 
($510 million or 84 percent) involved the 32 percent of the contracts 
from 1978 through 1982 that were still active (see fig. 3). Contracts 
issued from 1983 through 1997 accounted for the remaining $95 million.

Figure 3: Active v. Expired Contracts and Expended v. Unexpended 
Amounts from Contracts Issued from 1978 through 1982: 

[See PDF for image]

Note: Data are current as of June 2003.

[End of figure]

Based on average current spending rates calculated from AMAS data and 
projected forward using OMB inflation rates, RHS will not exhaust all 
the unexpended balances from these contracts until at least 2011. In 
2002, approximately $179 million in rental assistance funds was paid to 
project owners from contracts issued from 1978 through 1997, $53 
million of it from contracts issued from 1978 through 1982, and $126 
million from contracts issued from 1983 through 1997. At this rate, 
contracts from the 1983 to 1997 period will likely expend their 
remaining $95 million during 2004. The 1978 to 1982 contracts, which 
were funded based on inflation projections of 10 to 20 percent, will 
not expend their $510 million in unexpended balances until 2011 on 
average--8 years after the last of the 20-year contracts should have 
expired.

Figure 4: Projected Expenditures from Contracts Issued from 1978 
through 1997: 

[See PDF for image]

Note: Contracts issued from 1978 through 1982 include 20-and 5-year 
contracts.

[End of figure]

USDA Stated that Deobligating Unexpended Balances Would Result in 
Breach of Contract: 

The USDA regulations state that rental assistance contracts are 
effective for, depending on the contract, 5 or 20 years from the 
effective date of the agreement. These same regulations, however, make 
it clear that the expiration date of a contract is at complete 
disbursement of the funds obligated to the contract. This date, as the 
USDA regulations state, may be "before or after" the 5-or 20-year term.

The rental assistance contracts that implement this policy explicitly 
tie their expiration to the disbursement of rental assistance amounts 
listed in the contracts. In practice, this has resulted in many of the 
contracts extending beyond (in some instances, far beyond) the 
contemplated 5-or 20-year term. According to USDA, any effort to 
recapture the remaining unexpended funds associated with rental 
assistance agreements entered into from 1978 through 1982 would result 
in a breach of those contracts and would subject USDA to liability.

The Activity of Contracts Issued in 1998 through 2002 Is Consistent 
with Earlier Years: 

Ninety percent of the contracts issued from 1998 through 2002 are still 
active and appear to be expending their funds at a slower rate than RHS 
anticipated. Based on current expenditure rates, these contracts likely 
will run an average of over 6 years each. These findings are consistent 
with our analysis of RHS data, which shows that RHS has overestimated 
its spending needs most years since 1990.

The Majority of Contracts Issued from 1998 through 2002 Are Still 
Active and Expending Their Funds at a Slower Rate than RHS Anticipated: 

According to RHS data, and illustrated in figure 5, a relatively small 
percentage of contracts have expired. As of June 2003, 74 percent of 
the contracts issued in 1998 were still active, although the average 
contract should have expired by this date.[Footnote 9] The fact that so 
many were still active suggests that the majority of the 1998 contracts 
may have been overfunded. Furthermore, about 25 percent of the funds 
remained from the contracts issued in 1998, and about 35 percent of the 
funds remained from the contracts issued in 1999; only 11 percent of 
the funds allocated in 2002 were spent during the contracts' first 1½ 
years. This suggests that the contracts are also expending their funds 
more slowly than the 5 years on which RHS bases its budget needs.

Figure 5: Active v. Expired Contracts and Expended v. Unexpended 
Amounts from Contracts Issued from 1998 through 2002: 

[See PDF for image]

Note: Data are current as of June 2003.

[End of figure]

A document provided by the agency indicates that, since 1992, contracts 
have been spending on average 3 percent in their first year, 14 to 18 
percent in the second through sixth years, 8 percent in the seventh 
year, and tapering off around the twelfth year. According to the 
agency, this tapering indicates that overfunding of contracts is 
moderate. However, it should be noted that the document projects, for 
example, that about 1.7 percent of the funds allocated in 2000 will 
remain by the tenth year--a balance of $10.8 million from the $640 
million originally allocated.

Using RHS rental assistance payment data, we calculated that RHS 
overestimated its funding needs for these contracts by an average of 
about 8 percent each year. Between 1998 and 2002, almost $1.2 billion 
in rental assistance payments were made from contracts originating in 
those years, at an average rate of $2,808 per-unit per-year. However, 
RHS budgeted these units at an average annual rate of $3,019--a 
difference of $211 per-unit per-year (7.5 percent), or $1,055 per-unit 
per 5-year contract.

RHS Has Overestimated Its Spending Needs in Most Years Since 1990: 

Our analysis of rental assistance payment data showed that the agency 
has been overestimating its budget needs since at least 1990, the 
earliest year for which we gathered data. Importantly, where we had 
sufficient data from the agency, our analysis also shows that if RHS 
had used and correctly applied OMB inflation rates to its base per-unit 
rates, its estimates would have been closer to actual expenditures.

Figure 6 below provides an example of the difference between RHS's 
actual and estimated expenditures. The actual expenditures are averaged 
from the entire portfolio of 5-year contracts issued from 1989 through 
2002, while the estimated expenditures are averaged from only those 
units for which renewal, new construction, or servicing contracts 
originated in the corresponding year. Furthermore, the RHS estimated 
expenditure for a given year shows the effect of the 5-year estimate in 
the first year only. Due to RHS's method for calculating its estimated 
expenditures over a 5-year period, the difference is largest in the 
first year and declines over time as inflation raises the actual 
expenditure (or more accurate estimation) closer to the estimated 
expenditure (see fig. 2). The declining differentials of the second to 
fifth years are not reflected in figure 6 below. Nonetheless, while the 
estimated expenditures for any given year represent about 20 percent of 
the portfolio, they represent almost the entire portfolio over any 5-
year period in the figure. RHS estimates are above actual expenditures 
in each of the years. In addition, the corresponding estimated 
expenditures using OMB inflation rates also helps to illustrate the 
degree to which the RHS method has lead to overestimation. Sufficient 
information was not available from the agency to extend our OMB-based 
estimate of RHS expenditure prior to 2000. Our scope and methodology 
section contains a discussion of the data limitations we faced in our 
assessment of activity levels.

Figure 6: Actual v. Estimated Rental Assistance Expenditures Per-Unit, 
Per-Year, from 1990 through 2003: 

[See PDF for image]

Note: (1) RHS estimated expenditures for 1990 to 2002 are based on the 
agency's allocation estimates prior to obligation of appropriated 
funds; complete data of original estimates for the President's Budget 
were not available from the agency. As noted, in some years RHS's 
allocation amounts were smaller than the original submitted budgets; 
therefore, the RHS estimated numbers depicted may be smaller than the 
per-unit per-year amounts the agency actually received through 
appropriations. The 2003 estimated expenditures are based on the 
request submitted for the President's Budget. (2) RHS estimated 
expenditures show the effect of the 5-year estimate in the first year. 
Over time, the difference will decline as inflation raises the actual 
expenditures closer to the estimated expenditures. (3) Actual 
expenditures are for 5-year contracts issued 1989 to 2002; 2003 actual 
expenditures are estimated based on 10 months of available data. (4) 
RHS did not keep documentation of the inflation rates it used prior to 
2000; therefore, we did not have sufficient information to estimate the 
impact of using OMB rates in its estimates prior to that time.

[End of figure]

Between 1998 and 2002, RHS planned to fund 5-year contracts for an 
average of 42,000 units each year; the difference between the actual 
rate of expenditure and what RHS budgeted may mean that the agency had 
a surplus of approximately $43 million per year during this period, or 
more than $220 million total over the last 5 years.

Based on current spending rates, and allowing for inflation, the 
average contract issued during these years will likely run out of funds 
during its sixth year. That is, the average contract issued in 2000 or 
2001 will completely expend its funds during 2006 or 2007, 
respectively, and the average contract issued in 2002 will completely 
expend its funds during 2008.

Conclusions: 

RHS provides subsidized rental housing to almost half a million people 
each year. The rental assistance program, with an annual budget of over 
$700 million, provides further subsidies to about half this number. RHS 
budget estimating processes have caused the agency to overstate its 
spending needs over the life of the rental assistance program, 
resulting in hundreds of millions of dollars in unexpended balances. 
Consistently overstating funding needs for one program also undermines 
the congressional budget process by not allowing those funds to be 
available for other programs. RHS is updating and automating its budget 
estimation process, and its new forecasting model shows some 
improvements over past and current processes. However, there are some 
flaws with the model. The agency plans to use its own inflation rates, 
which are based on historic rates of change, rather than the inflation 
rates provided by OMB, which predict future rates of change. 
Furthermore, RHS is incorrectly calculating the rates it plans to use, 
which may cause the agency to underestimate its future contract needs. 
A simple modification to the agency's planned budget estimation process 
would help the agency more accurately estimate its rental assistance 
needs and curtail future unexpended balances--or budget shortfalls as 
the case may be.

Recommendation for Executive Action: 

To more accurately estimate rental assistance budget needs, we 
recommend that the Secretary of Agriculture require program officials 
to use and correctly apply the inflation rates provided by OMB in its 
annual budget and allocation estimation processes.

Agency Comments and Our Evaluation: 

We provided USDA and OMB with a draft of this report for their review 
and comment. The Acting Undersecretary for Rural Development for USDA 
raised several concerns about our analysis of RHS's budgeting practices 
and rental assistance expenditure data. In particular, the Acting 
Undersecretary argued that OMB's Circular A-11 encourages the use of a 
per-property microscale rate of change, rather than a blanket national 
rate. OMB's Circular A-11 states that all budget materials must be 
consistent with the economic assumptions provided by OMB. While OMB 
policy permits consideration of certain factors in developing out-year 
estimates, this does not mean that an agency should automatically use 
its own economic assumptions without providing documentation and 
justification. As we note in our report, neither OMB staff nor the 
Rural Development Budget Office required a justification for the 
agency's inflation rates.

Furthermore, according to the Acting Undersecretary, we did not 
demonstrate that using inflation rate projections from the President's 
Budget would provide a more accurate budget estimate. We believe that 
figure 6 in this report illustrates that using the inflation 
projections from the President's Budget would have brought the agency's 
budget estimates closer to its actual expenditures, without running the 
risk of underfunding the rental assistance contracts.

USDA also disagreed with our finding that RHS's budget estimates were 
too high. As stated in this report, we believe RHS overestimates its 
budget needs by using inflation factors that are higher than those 
projected by OMB for use in the budget process, improperly compounding 
the inflation rates, and using expenditure rates that may overstate the 
need for rental assistance. Our estimates reflect the extent to which 
RHS may have overestimated its budgets when compared with estimates 
based on OMB budget guidance documents and appropriate application of a 
compounding formula. USDA's complete written comments and our responses 
appear in appendix I.

We received oral comments from OMB. OMB representatives did not have 
comments on the specific findings or conclusions of this report. They 
did, however, note that they are always open to suggestions that would 
help the Administration provide more accurate budget projections along 
with the related oversight.

Scope and Methodology: 

To assess the accuracy of RHS's budget estimates for the rental 
assistance program, we collected written and testimonial information 
from agency officials on current budget estimation methods and the 
budget automation plan that is being developed. We reviewed OMB 
guidance on preparing agency budgets and the inflation rates provided 
by OMB for agency use, and interviewed OMB staff that oversee the 
rental assistance program. Finally, we consulted our Standards for 
Internal Control in the Federal Government to review control activities 
that apply to RHS's budget estimation processes. In describing RHS's 
current process for estimating its budget needs, we faced the problem 
that the agency has no official written documentation for that process. 
Most information was provided verbally, and the information the agency 
did provide outlined only its elementary budget processes.

To assess the activity level of rental assistance contracts issued from 
1978 through 1997 with unexpended balances, we reviewed rental 
assistance data from the agency's Automated Multi-Housing Accounting 
System (AMAS) from January 1990 through October 2003 to determine the 
extent of the unexpended balances. We also used these data to determine 
the rate at which those balances were currently being expended; by 
applying OMB inflation rates for future years to the current rates of 
expenditure, we estimated when the funds will expire. We acquired OMB 
inflation rates for future years from the fiscal year 2004 and 2005 
President's Budgets.[Footnote 10]

To assess the activity level of rental assistance contracts issued from 
1998 through 2002 and the accuracy of RHS's estimates of the rates at 
which these funds would be used, we reviewed rental assistance data 
from AMAS from January 1998 through October 2003 to determine the 
activity level of the unexpended balances. We also used these data to 
determine the rate at which those balances were currently being 
expended; by applying OMB inflation rates for future years to current 
rates of expenditure we estimated when the funds will expire. We 
assessed the accuracy of RHS's estimates of the rate at which the funds 
would be used by comparing RHS's estimated rental assistance 
expenditures to actual program expenditures. We determined RHS's 
estimated expenditures based on data provided by the agency. We 
determined actual program expenditures using payment data from AMAS.

We faced limitations in our assessment of actual program expenditures 
using AMAS data. We had compared actual expenditures averaged from the 
entire portfolio of contracts issued from 1978 through 2002 with 
estimated expenditures averaged from only those units for which 
renewal, new construction, or servicing contracts originated in the 
corresponding year. In response to agency concerns, we eliminated 
contracts issued prior to 1989 because they represented a mix of 
contracts that were expending funds normally and contracts that 
exhibited unusual behavior resulting in abnormally low expenditures. 
The agency opined that the annual expenditures from these abnormal 
contracts could not be compared with their estimates. Due to the 
structure of AMAS data, we were unable to isolate the abnormal behaving 
contracts. However, the resulting figure (fig. 6), based on their 
comments, looks very similar to the original figure.

For the AMAS data we used, we requested and received the most current 
data available from the system. We assessed the reliability of the data 
by (1) reviewing existing information about the systems and the data, 
(2) interviewing agency officials knowledgeable about the data, and (3) 
examining the data elements (fields) used in our work by comparing 
known and/or anticipated values. When inconsistencies were found, we 
discussed our findings with agency officials to understand why 
inconsistencies could exist. We determined that the data were 
sufficiently reliable for the purposes of this report.

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 also 
will 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  [Hyperlink, 
http://www.gao.gov].

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 III.

Sincerely yours,

Signed by: 

William B. Shear: 
Director, Financial Markets and Community Investment: 

[End of section]

Appendixes: 

Appendix I: Comments from the U.S. Department of Agriculture: 

USDA:

United States Department of Agriculture:

Office of the Secretary 
Washington, D.C. 20250:

MAR 8 2004:

Mr. William B. Shear:

Director, Financial Markets and Community Investment: 
United States General Accounting Office:

441 G Street, NW Room 2A10: 
Washington, DC 20548:

Dear Mr. Shear,

Thank you for providing the United States Department of Agriculture 
(USDA), Rural Development, with your draft report on Standardization of 
Budget Estimation Processes Needed for Rental Assistance Program, Audit 
No. GAO-04-424. I would like to offer the following comments for your 
consideration and ask that a copy of this response be included in your 
final report.

USDA provides affordable multifamily housing to almost a half-million 
residents in rural America. One of the most important tools available 
is the Rental Assistance (RA) program, which enables eligible very low 
and low-income residents to pay only 30 percent of their income towards 
rent. USDA takes very seriously its stewardship of this program.

While USDA acknowledges that there is evidence that we have 
inadvertently overestimated the cost of RA contracts in the past, we do 
not believe that the report supports an estimation of $51 million a 
year in over funding in Fiscal Year (FY) 2003. It should be noted that 
at no point were government funds at risk and there has been no loss of 
funds due to the Agency's RA budget estimation policy. Funds 
appropriated for RA contracts have allowed any balance remaining at the 
end of the five-year contract period to be used for the extension of 
those contracts. Otherwise, these contracts would have required 
additional funds for renewal.

The General Accounting Office (GAO) found that contracts are now 
expected to exhaust funds in the sixth year. USDA believes this 
validates the method of estimating that Rural Housing Service (RHS) has 
used in recent years. RHS has succeeded in narrowing down estimated 
expenditure rates without the benefit of complete automation. RHS's:

estimating methods have evolved. Over time, assumptions, factors and 
circumstances require us to re-evaluate and change our methods. 
Multifamily property and health insurance costs, taxes and operating 
expenses all increase over time. Given the dynamic nature of our 
portfolio, we believe it is inconclusive to determine at this time that 
active contracts within their renewal period are over funded.

GAO identified three ways that overestimating occurs: inflation rate, 
compounding rate and expenditure rate.

Inflation Rate:

GAO notes that RHS has not used the Office of Management and Budget 
(OMB) inflation rate or followed OMB procedures. However, USDA would 
argue that OMB Circular A-1 l actually encourages the use of a per 
property micro-scale rate of change rather than a blanket national 
rate. We believe our budgeting is in accordance with OMB budget 
planning guidance as required by OMB Circular A-11.

OMB is currently consulted but we will implement a more formal process. 
As in the past, USDA will work with OMB to develop a mutually 
acceptable rate. USDA will also continue to evaluate use of the per 
project rate of change as an indicator of future RA use because we 
believe it is the best method of determining what is happening in a 
specific property.

GAO attempts to show in Figure 6 that RHS estimates have exceeded 
actual expenditures for 1990-2003. However, Figure 6 is flawed for two 
reasons: (1) it presents only Year 1 of a five-year contract to compare 
estimates to expenditures. A truer picture of any overestimation claim 
would follow a contract through its life span and compare the estimate 
against actual expenditures. Showing only the first year gives a 
distorted view, and (2) as noted by GAO, RHS budget estimation process 
data was not available for years prior to 2000. We were unable to trace 
the per unit values shown back to any source.

Inflation Rate Compounding Calculation:

USDA agrees that the compounding calculation used most recently may 
have resulted in contracts outliving their expected renewal term by 
three to six months. Similarly, the proposed RA Forecasting Tool has 
been revised to address GAO's preliminary tool review concerns.

Expenditure Rate:

The third factor cited by GAO for overestimation of the RA budget needs 
is the expenditure rate that RHS used in its calculations. It is 
important to note that RHS developed the average rate in an effort to 
balance out the high and low vacancy rates; vacancies were not ignored 
as claimed in the report. It should also be noted that RA is only paid 
on occupied units and not on vacant units. USDA believes that using the 
most recent 3 months' history represented a more realistic picture of 
occupancy going forward vs. the occupancy 12 months previous, absent 
the availability of historic property data, now available with the 
proposed automated Forecasting Tool. Also through this process, RHS 
instilled quality control by investigating any abnormally high or low 
rates prior to calculating the contract amount; we did not 
automatically renew without examination.

USDA believes the discussion in the report does not convey RHS's 
attempts to achieve balance in its estimating calculations.

USDA reiterates that GAO has not demonstrated that use of OMB's 
inflation rate would result in exact five-year contracts or that use of 
the OMB rate matches the RA needs at our multifamily properties. To 
require use of a blanket national rate as a basis for funding RA 
contracts jeopardizes the property and the tenants. Under funding 
contracts is as equally serious a problem as over funding. One of the 
major reasons RHS developed the proposed automated RA Forecasting Tool 
is to have the ability to drill down to a property level and determine 
that specific property's RA needs. USDA is committed to protecting 
tenants and the financial viability of the housing they occupy.

USDA disagrees with GAO's assertion that the 1998-2003 contracts are 
expending funds at a level consistent with contracts funded in 1978-
1997. The following chart reflects that the contracts funded in the 
last six years (1998-2003) are expending their funds at a faster rate 
than the historic rate.

Expenditure Rates FY 1998 - 2003 (Historic vs. Current):

[See PDF for image]

Note: Year I represents FY 2003 contracts in their I ° year of renewal; 
Year 2 represents FY 2002 contracts in their 2'D year of renewal, etc. 
through FY 1998 contracts in 6` year). Data as of March 4, 2004.

[End of figure]

RHS, through a number of methods, monitors the appropriate use of RA, 
guards against fraud, waste and abuse, and continues to improve and 
revamp its procedures and processes. USDA believes it would be an error 
to base any budgetary decisions on the claim of $51 million in over 
funding for FY 2003 simply because we did not use the OMB inflation 
rate. FY 2003 contracts have just started their five-year renewal 
cycle.

USDA is committed to the future of rural communities and we continue to 
work toward improving our rental assistance program.

Thank you for this opportunity to comment on the report. If you have 
any questions, please contact John M. Purcell, Director, Financial 
Management Division, at (202) 692-0080.

Signed by: 

GILBERT G. GONZALEZ: 
Acting Under Secretary Rural Development: 

The following are GAO's comments on the U.S. Department of 
Agriculture's letter dated March 8, 2004.

GAO Comments: 

1. Our $51 million figure is based on data provided by the U.S. 
Department of Agriculture (USDA), analyzed in a manner consistent with 
the Office of Management and Budget (OMB) budgetary guidance, 
correcting for the areas where we believe USDA overestimates its 
budgetary requirements. As we state in the report, this number is an 
estimate of the extent to which the Rural Housing Service (RHS) may 
have overestimated its budget when compared with an estimate based on 
OMB budget guidance documents and appropriate application of a 
compounding formula.

2. Although we concur that a budget estimation process, by itself, will 
not necessarily put government funds at risk, consistently overstating 
funding needs undermines the congressional budget process. In addition, 
without performing a detailed internal controls review, we cannot state 
that the current process for allocating budget funds has not put 
government funds at risk or led to a loss of funds. We do note that RHS 
is not adhering to internal control standards regarding segregation of 
duties for both its estimation and allocation processes, and such an 
internal control lapse could introduce a risk of error or fraud.

3. We agree that RHS's contracts are not lasting as long as they did in 
the past, however, a 6-year average life contract is still 20 percent 
greater than the intended contract life.

4. Circular A-11 states that "all budget materials, including those for 
out-year policy and baseline estimates, must be consistent with the 
economic assumptions provided by OMB. OMB policy permits consideration 
of price changes for goods and services as a factor in developing 
[outyear] estimates. However, this does not mean that you should 
automatically include an allowance for the full rate of anticipated 
inflation in your request." If the agency has evidence that a property 
will perform above or below the Consumer Price Index, which is the 
basis for OMB's economic assumptions, we would agree that this evidence 
should be used. OMB guidance indicates that the agency should document 
this evidence and justify that its proposed budget-estimation 
methodology would create a more accurate budget estimate.

5. We note in our report that the difference declines over time. Figure 
2 shows this decline and shows that RHS is still overestimating its 
budget needs in the fifth year, albeit by less than in years 1 through 
4.

6. As stated in the note in figure 6, RHS's estimated expenditures are 
based on data provided by the agency. Actual RHS expenditures are based 
on data from RHS's Automated Multi-Housing Accounting System (AMAS). As 
noted in the report, RHS did not document the inflation rate it used in 
its budget and allocation estimates prior to 2000. This is the lack of 
data to which we refer. It only affected our OMB-based estimate of RHS 
expenditures by preventing us from backing out agency rates and 
replacing them with the inflation projections from the President's 
Budget for years prior to fiscal year 2000. We clarified the text on 
this point.

7. Our concern centers on the fact that the agency is using the highest 
of the most recent 3 months instead of an average of all 3 months, not 
on the use of 3 months of data versus 12 months of data.

8. We concur that these rates are, in fact, an estimate, but as figure 
6 illustrates, using these rates would have brought the agency's 
estimates closer to its actual expenditures. Furthermore, as shown in 
figure 6, budget estimates based on the inflation projections from the 
President's Budget would still have been higher than actual program 
expenditures, which should alleviate USDA's concern that using this 
method would underfund contracts.

9. Our report does not make this assertion. We state that the activity 
of contracts issued from 1998 through 2002 is consistent with earlier 
years, and in particular, that RHS has overestimated its spending needs 
in most years since 1990.

10. The $51 million overestimate does not stem from the inflation rate 
only. It also stems from RHS compounding the rate to the fifth power 
and then applying that rate back to each year of the contract. We 
concur that we will not know the outcome of the contracts issued in 
2003 until 2008 or later. This is our best estimate based on data 
provided by the agency and following OMB budgetary guidance.

[End of section]

Appendix II: GAO Contacts and Staff Acknowledgments: 

GAO Contacts: 

Andy Finkel (202) 512-6765 f [Hyperlink, finkela@gao.gov] 
inkela@gao.gov Katherine Trimble (202) 512-5033 t [Hyperlink, 
trimblemk@gao.gov] rimblemk@gao.gov: 

Staff Acknowledgments: 

In addition to those named above, William Bates, Emily Chalmers, Jamila 
Jones, Austin Kelly, Marc Molino, and Julie Trinder made key 
contributions to this report.

(250122): 

FOOTNOTES

[1] An agency's cancellation or downward adjustment of previously 
recorded obligations.

[2] We will issue a separate legal opinion on this issue.

[3] The Section 515 Multifamily Direct Rural Rental Housing Loans 
program provides loans, subsidized as low as 1 percent interest, to 
developers to build rental housing. The Section 514 Multifamily Housing 
Farm Labor Loans program provides grants and loans subsidized at 1 
percent interest 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 projects must be established on a 
nonprofit or limited profit basis to receive rental assistance.

[4] Eligible 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 
(AMI); low income is between 50 and 80 percent of AMI. 

[5] The study did not consider 20-year contracts, likely because the 
first of these would not expire until 1998.

[6] The actual inflation rate, as measured by the Consumer Price Index, 
ranged from 8.9 percent to 13.3 percent from 1978 through 1981, and 
fell to 3.8 percent in 1982.

[7] In 2002, for example, there was a difference of $30 million 
(overestimation) between the agency's budget request and actual 
allocation needs.

[8] U.S. General Accounting Office, Standards for Internal Control in 
the Federal Government, GAO/AIMD-00-21.3.1 (Washington, D.C.: Nov. 1, 
1999).

[9] Contracts are funded quarterly, based on their estimated renewal 
needs between January and December. Therefore, the average contract is 
funded based on a June renewal time frame, and an average contract 
lifetime would run from June of the first year to June of the fifth 
year (e.g., June 1998 to June 2003).

[10] Table S-11: Comparison of Economic Assumptions from the Budget for 
Fiscal Year 2005, lists inflation rates for 2005 to 2009. For any 
projections 2010 and beyond, we used the rate for 2009.

GAO's Mission: 

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

Obtaining Copies of GAO Reports and Testimony: 

The fastest and easiest way to obtain copies of GAO documents at no 
cost is through the Internet. GAO's Web site ( www.gao.gov ) contains 
abstracts and full-text files of current reports and testimony and an 
expanding archive of older products. The Web site features a search 
engine to help you locate documents using key words and phrases. You 
can print these documents in their entirety, including charts and other 
graphics.

Each day, GAO issues a list of newly released reports, testimony, and 
correspondence. GAO posts this list, known as "Today's Reports," on its 
Web site daily. The list contains links to the full-text document 
files. To have GAO e-mail this list to you every afternoon, go to 
www.gao.gov and select "Subscribe to e-mail alerts" under the "Order 
GAO Products" heading.

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. General Accounting 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, Waste, and Abuse in Federal Programs: 

Contact: 

Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: fraudnet@gao.gov

Automated answering system: (800) 424-5454 or (202) 512-7470: 

Public Affairs: 

Jeff Nelligan, managing director, NelliganJ@gao.gov (202) 512-4800 U.S.

General Accounting Office, 441 G Street NW, Room 7149 Washington, D.C.

20548: