Multifamily Housing: HUD Missed Opportunities To Reduce Costs on its
Uninsured Section 8 Portfolio (Chapter Report, 07/30/1999,
GAO/RCED-99-217).

Pursuant to a congressional request, GAO reviewed the Department of
Housing and Urban Development's (HUD) portfolio of section 8 properties
with that are not insured by the Federal Housing Administration,
focusing on: (1) section 8 rental assistance provided to properties in
HUD's uninsured portfolio; (2) the financial benefits that may be
available to state and local housing agencies that participate in the
section 8 program; and (3) the ways HUD and the state agencies oversee
the physical and financial condition of the properties in their
respective uninsured section 8 portfolios and the information they have
on the physical and financial condition of these properties.

GAO noted that: (1) according to HUD's data, rental assistance payments
for the uninsured section 8 portfolio totalled over $3.3 billion in
fiscal year 1998; (2) a majority of these payments--about $2.3
billion--were associated with the two largest programs in the uninsured
portfolio, one of which is the state agency program; (3) complete data
were not available for assessing the relationship of the rents for
section 8 units, or Section 8 contract rents, to actual market rents;
(4) nevertheless, available information indicates that some contract
rents exceed market rents; (5) when section 8 contract rents exceed
market rents, the section 8 subsidies support higher rents than the
properties generally command without federal assistance; (6) the federal
government will continue to incur these high costs each year until its
existing section 8 contracts expired; (7) these contracts will expire at
various times, generally from within the next 5 years to about 20 years;
(8) contracts in the state agency program will generally be among the
last to expire; (9) under the section 8 program, state and local
agencies may derive financial benefits, or savings, from refunding their
tax-exempt bonds; (10) the agencies may receive one of two available
fees for administering their section 8 contracts; (11) the agencies are
required to use the savings from refunding their bonds, and in some
cases may use a portion of their fee, to provide affordable housing for
low-income residents within their jurisdictions; (12) GAO found that HUD
has not resolved three long-standing issues associated with these
financial benefits; (13) as a result, HUD has missed opportunities to
reduce its section 8 costs by tens of millions of dollars, particularly
in the state agency program; (14) to monitor the physical and financial
condition of properties in the uninsured portfolio, HUD requires annual
physical inspections and generally requires annual audited financial
statements; (15) as of December 1998, HUD had limited information on the
physical condition of properties in the uninsured portfolio and no
information on their financial condition; (16) according to HUD's
central database, which included the results of inspections for about 63
percent of the properties, most of the properties were in satisfactory
or better physical condition, however, these ratings were not based on
objective criteria and their reliability is therefore unknown; and (17)
in mid-1998, HUD began to establish centralized procedures to improve
the monitoring of multifamily properties in its uninsured and other
portfolios.

--------------------------- Indexing Terms -----------------------------

 REPORTNUM:  RCED-99-217
     TITLE:  Multifamily Housing: HUD Missed Opportunities To Reduce
	     Costs on its Uninsured Section 8 Portfolio
      DATE:  07/30/1999
   SUBJECT:  Housing programs
	     Rent subsidies
	     Rental housing
	     Low income housing
	     State-administered programs
	     Federal/state relations
	     Cost control
	     Rental rates
IDENTIFIER:  HUD Section 8 Housing Assistance Program

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed 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.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************

Cover
================================================================ COVER

Report to Congressional Committees

July 1999

MULTIFAMILY HOUSING - HUD MISSED
OPPORTUNITIES TO REDUCE COSTS ON
ITS UNINSURED SECTION 8 PORTFOLIO

GAO/RCED-99-217

HUD's Uninsured Section 8 Portfolio

(385713)

Abbreviations
=============================================================== ABBREV

  FHA - Federal Housing Administration
  HUD - Department of Housing and Urban Development
  OIG - Office of the Inspector General

Letter
=============================================================== LETTER

B-281431

July 30, 1999

Letter
=============================================================== LETTER

Congressional Committees

This report was prepared to comply with the requirements of section
532 of the 1998 Appropriations Act for the Departments of Veterans
Affairs and Housing and Urban Development, and Independent Agencies
(P.L.  105-65, Oct.  27, 1997), which requires a GAO study of HUD's
portfolio of properties with Section 8 project-based rental
assistance that are not insured by the Federal Housing
Administration.  As agreed, this report provides information on the
Section 8 rental assistance provided to properties in HUD's uninsured
Section 8 portfolio, the benefits that may be available to state and
local housing agencies under the Section 8 program, and the ways HUD
and the state agencies oversee the physical and financial condition
of the properties in their respective uninsured Section 8
project-based portfolios and the information they have on the
physical and financial condition of these properties. 

We are sending copies of this report to congressional committees and
subcommittees interested in housing, the Secretary of Housing and
Urban Development, the Director of the Office of Management and
Budget, and other interested parties.  We will also make copies
available to others upon request. 

If you or your staff have any questions about this report, please
call me at (202) 512-7631.  Key contributors to this report are
listed in appendix VIII. 

Judy A.  England-Joseph
Director, Housing and Community
 Development Issues

Enclosure

List of Committees

The Honorable Wayne Allard
Chairman, Subcommittee on Housing and Transportation
Committee on Banking, Housing and Urban Affairs
United States Senate

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

The Honorable Christopher S.  Bond
Chairman, Subcommittee on VA, HUD and
 Independent Agencies
Committee on Appropriations
United States Senate

The Honorable Barbara A.  Mikulski
Ranking Minority Member, Subcommittee on
 VA, HUD and Independent Agencies
Committee on Appropriations
United States Senate

The Honorable Rick Lazio
Chairman, Subcommittee on Housing
 and Community Opportunity
Committee on Banking and Financial Services
House of Representatives

The Honorable Barney Frank
Ranking Minority Member, Subcommittee on
 Housing and Community Opportunity
Committee on Banking and Financial Services
House of Representatives

The Honorable James T.  Walsh
Chairman, Subcommittee on VA, HUD
 and Independent Agencies
Committee on Appropriations
House of Representatives

The Honorable Alan B.  Mollohan
Ranking Minority Member, Subcommittee on
 VA, HUD and Independent Agencies
Committee on Appropriations
House of Representatives

EXECUTIVE SUMMARY
============================================================ Chapter 0

   PURPOSE
---------------------------------------------------------- Chapter 0:1

Section 8 rental housing assistance is the main form of federal
housing assistance for the nation's low-income residents.  Section 8
assistance is tied either to units in specific properties
(project-based assistance) or to families and individuals who live in
affordable rental housing of their choice (tenant-based assistance). 
The residents of housing units that receive project-based assistance
generally pay 30 percent of their income for rent, while the
Department of Housing and Urban Development (HUD) pays the balance. 
HUD provides Section 8 project-based rental assistance to units in
approximately 22,000 multifamily properties, almost half of which
receive mortgage insurance from HUD's Federal Housing Administration
(FHA).  The remaining properties, which are not insured by FHA, are
referred to collectively as HUD's ï¿½uninsuredï¿½ Section 8 portfolio. 

A mandate for a study of the uninsured portfolio was included in
HUD's fiscal year 1998 appropriations bill (P.L.  105-65, Oct.  27,
1997).  Accordingly, this report examines (1) the information HUD has
on the Section 8 assistance provided to properties in the uninsured
portfolio, (2) the financial benefits that may be available to state
and local housing finance agencies that participate in the Section 8
program and the impact of these benefits on the Section 8 program's
costs, and (3) the ways HUD and the state agencies oversee the
physical and financial condition of the properties in their
respective uninsured Section 8 portfolios and the information they
have on the physical and financial condition of these properties. 

   BACKGROUND
---------------------------------------------------------- Chapter 0:2

As of December 1998, according to HUD's data, the uninsured Section 8
portfolio consisted of 12,708 contracts between HUD and property
owners.  These contracts cover 632,216 assisted units associated with
eight programs, including a state agency program.  Most of these
programs were established in the 1970s to develop housing for
low-income households, using various types of financing and long-term
(20- to 40-year) Section 8 contracts.  While some of the properties
were financed by loans and grants from HUD, others were financed by
bonds issued by state and local housing finance agencies (state and
local agencies).  All but one of the housing development programs
were terminated in 1983 because of high costs, but many of the
Section 8 contracts for properties developed through these programs
are still in effect.  HUD will continue to incur rental assistance
costs until these contracts expire. 

During the late 1970s and early 1980s, the cost of bonds to finance
housing development rose with interest rates to unprecedented levels. 
HUD therefore authorized higher mortgage interest rates and higher
rental assistance payments to cover the higher bond financing costs,
first in 1980 and then in 1981.  Only the Section 8 contracts covered
by the 1981 authorization required the agencies to refund (refinance)
their bonds when interest rates declined and to provide all of the
savings (the difference between the original and the current debt
service costs) to the federal government. 

The Stewart B.  McKinney Homeless Assistance Amendments Act of 1988,
enacted primarily to assist the nation's homeless, included one
section (1012) that provided for the state agencies to share the bond
refunding savings that they were formerly required to return to the
federal government and to use these savings to provide affordable
housing for households with very low incomes.  While section 1012
originally applied only to refundings associated with Section 8
contracts covered by HUD's 1981 authorization, an October 1992
amendment to section 1012 apparently provided for sharing the savings
from refunding bonds associated with other Section 8
contractsï¿½savings that some state agencies were generally accustomed
to retaining. 

In 1992 and 1993, HUD's Office of the Inspector General released two
reports that examined whether bond financed Section 8 properties were
refinanced as intended and if HUD realized the appropriate savings
from the bond refundings.\1 The reports disclosed, among other
things, that HUD had not fully realized potential savings from bond
refundings and identified actions that HUD could take to realize
additional savings. 

--------------------
\1 Interim Audit Report Bond Refundings of Section 8 Projects, Office
of the Inspector General (93-HQ-119-0004, Oct.  30, 1992) and
Multi-Region Audit of Refunding of Bonds for Section 8 Assisted
Projects, Office of the Inspector General (93-HQ-119-0013, Apr.  30,
1993). 

   RESULTS IN BRIEF
---------------------------------------------------------- Chapter 0:3

According to HUD's data, rental assistance payments for the uninsured
Section 8 portfolio totaled over $3.3 billion in fiscal year 1998.  A
majority of these paymentsï¿½about $2.3 billion--were associated with
the two largest programs in the uninsured portfolio, one of which is
the state agency program.  Complete data were not available for
assessing the relationship of the rents for Section 8 units, or
Section 8 contract rents, to actual market rents.  Nevertheless,
available information indicates that some contract rents exceed
market rents.  When Section 8 contract rents exceed market rents, the
Section 8 subsidies support higher rents than the properties
generally could command without federal assistance.  Moreover, the
federal government will continue to incur these high costs each year
until its existing Section 8 contracts expire.  These contracts will
expire at various times, generally from within the next 5 years to
about 20 years.  Contracts in the state agency program will generally
be among the last to expire. 

Under the Section 8 program, state and local agencies may derive
financial benefits, or savings, from refunding their tax-exempt
bonds.  In addition, the agencies may receive one of two available
fees for administering their Section 8 contracts.  The agencies are
required to use the savings from refunding their bonds, and in some
cases may use a portion of their fee, to provide affordable housing
for low-income residents within their jurisdictions.  GAO found that
HUD has not resolved three long-standing issues associated with these
financial benefits.  As a result, HUD has missed opportunities to
reduce its Section 8 costs by tens of millions of dollars,
particularly in the state agency program.  First, HUD has not issued
guidance to the state agencies on how to comply with the October 1992
amendment to section 1012 of the McKinney Act, which provides for the
agencies to share certain bond refunding savings with the federal
government.  As a result, some state agencies have retained all of
the savings, which accrue annually, while other agencies have shared
the savings.  Second, HUD has not provided clear guidance to the
state and local agencies for calculating rent increases after
refunding bonds.  Consequently, the Section 8 rental assistance
program is incurring excess costs that could have been avoided. 
Finally, HUD has allowed some state agencies to collect both of the
available fees for administering their Section 8 contracts, despite a
1980 HUD regulation prohibiting dual fees.  These fees, provided by
the Section 8 program, can amount to millions of dollars per year. 
HUD has known about these issues since at least 1992, when the
Inspector General first reported on them, but it has not acted
quickly or effectively to resolve them.  As a result, the federal
government has lost opportunities to share bond refunding savings and
has incurred excessive rental assistance and administrative fee
payments.  GAO makes several recommendations to HUD on each of these
issues to reduce the Section 8 costs borne by the government (see ch. 
3). 

To monitor the physical and financial condition of properties in the
uninsured portfolio, HUD requires annual physical inspections and
generally requires annual audited financial statements.  Ten state
agencies, which monitor about half of the properties in the state
agency program, told GAO that they had monitoring policies and
procedures in place that conformed to HUD's guidance.\2 As of
December 1998, HUD had limited information on the physical condition
of properties in the uninsured portfolio and no information on their
financial condition.  According to the Department's central database,
which included the results of inspections for about 63 percent of the
properties, most of the properties were in satisfactory or better
physical condition; however, these ratings were not based on
objective criteria and their reliability is therefore unknown.  The
10 state agencies told GAO that 95 percent of the properties in their
portfolios were in satisfactory or better physical condition. 
Moreover, according to these agencies, only a very small fraction of
their properties (under 4 percent) warranted special monitoring
because of financial or other problems.  In mid-1998, HUD began to
establish centralized procedures, including objective rating
criteria, to improve the monitoring of multifamily properties in its
uninsured and other portfolios. 

--------------------
\2 GAO surveyed 10 state agencies (California, Illinois, Maryland,
Massachusetts, Michigan, Minnesota, New Hampshire, Oregon, Tennessee,
and Wisconsin) to obtain information on the methods they used to
monitor their properties and on the physical and financial condition
of the properties.  GAO also visited 5 of the 10 agencies. 

   PRINCIPAL FINDINGS
---------------------------------------------------------- Chapter 0:4

      INFORMATION ON THE UNINSURED
      PORTFOLIO VARIES BY PROGRAM
-------------------------------------------------------- Chapter 0:4.1

About 68 percent of the $3.3 billion in rental assistance for the
uninsured portfolio went to two of eight programs--the
elderly/disabled loan program and the state agency program.  The
average per-unit costs for the eight programs varied widely,
primarily because of differences in the ways properties were
financed.  The per-unit costs and, to a lesser extent, tenants'
income levels determine the Section 8 rent subsidies that HUD must
pay for assisted units.  For several of the programs, including the
state agency program, HUD's subsidies tended to be high because (1)
the assisted rents were initially set above market levels to
encourage the production of affordable housing and (2) the formulas
for automatic rent increases (which, until recently, were provided
each year) tended to be generous, according to HUD.  The Congress now
prohibits automatic rent increases for properties whose rents exceed
the rent standardsï¿½called fair market rents--that HUD develops
annually for geographic locations, such as large metropolitan areas. 
GAO determined that the Section 8 contract rents for about 75 percent
of the assisted units in the uninsured portfolio exceeded HUD's fair
market rents.  HUD's fair market rents may not be the same as actual
market rents because they do not reflect the differences in market
value that may be found from one neighborhood to another within a
geographic location.  However, nationwide data on rents for
particular neighborhoods were not available for assessing the
relationship of Section 8 contract rents to market rents. 
Nevertheless, the history and design of the Section 8 project-based
program, together with information from two states and a study of 53
bond-financed Section 8 properties,\3 indicate that some Section 8
contract rents exceed market rents in the uninsured portfolio.  When
Section 8 contract rents exceed market rents, the Section 8 subsidies
support higher rents than the properties could command without
federal assistance.  Moreover, these high subsidy costs will continue
until the existing Section 8 contracts expire.  While many of the
contracts will be expiring in the next 5 years, those in the state
agency program will generally expire in 10 to 20 years. 

--------------------
\3 HUD's Local Multifamily Portfolio, John Nuveen & Co., Inc.  (July
1997). 

      HUD HAS MISSED OPPORTUNITIES
      TO REDUCE COSTS IN THE STATE
      AGENCY PROGRAM
-------------------------------------------------------- Chapter 0:4.2

HUD has not issued clear guidance to the state agencies on sharing
savings with the federal government when they have refunded bonds
associated with certain Section 8 contracts covered by the October
1992 amendment to section 1012 of the McKinney Act.  While HUD has
required the agencies to share savings when their Section 8 contracts
include explicit requirements for providing savings to the
government, it has not required the agencies to share when their
contracts do not include such requirements.  In April 1996, HUD tried
to require the state agencies to share their savings by publishing a
regulation that was intended to establish the applicability of the
October 1992 amendment to all refundings by the state agencies. 
However, HUD made a typographical error in the regulation, citing the
wrong paragraphs of the McKinney Act.  As a result, the regulation
did not have its intended effect.  Moreover, in the view of the
National Council of State Housing Agencies and some state agencies,
the amendment generally applies to state agencies only when their
Section 8 contracts specify that they are to provide bond refunding
savings to the government.  GAO visited five state agencies that
refunded nearly all of their bonds in the mid-1990s.  Three of these
agencies generally do not share bond refunding savings with the
federal government except when their contracts direct them to provide
the savings to the government.  The other two agencies share savings
with the government whether or not their contracts direct them to
provide the savings to the government. 

When bonds issued by state and local agencies have been refunded but
rents subsidized by the government under Section 8 contracts have not
been reduced to reflect the bond refunding savings, rent increases
based on HUD's general method for calculating increases will be
excessive.  Although the Inspector General recommended in 1992 that
HUD take action to prevent these excessive rent increases, HUD
initially disagreed with the recommendation, maintaining it did not
have the authority to limit Section 8 rent increases.  Then, in
August 1997, HUD issued a notice establishing procedures for
considering the savings when calculating rent increases for contracts
that provided for returning the savings to the government.  However,
the notice did not include a methodology for implementing the
procedures or an example of a calculation.  Furthermore, the notice
was issued for 1 year, and HUD did not renew it when it expired.  HUD
officials told GAO that renewing the notice should not have been
necessary because the procedures apply to rent increases over the
lives of the Section 8 contracts.  Two of the five state agencies
that GAO visited calculated rent increases subject to the notice
during the year the notice was in effect.  One agency complied with
it, and the other did not.  The agency that did not comply said that
HUD did not provide it with the methodology to perform the
calculation.  The agency also stated that the requirement was no
longer in effect. 

HUD compensates the state and local agencies for administering their
Section 8 project-based rental assistance contracts.  The state
agencies are entitled to receive either an annual contributions
contract feeï¿½a per-unit fee provided by HUD--or an override feeï¿½a fee
that represents the difference between an agency's borrowing (bond
issuance) and lending rates.\4

The agencies are not allowed to receive both fees, according to a
Section 8 regulation promulgated in 1980.  Nevertheless, in 1992 the
Inspector General found, in reviewing the refunding of bonds
associated with Section 8 contracts, that some state agencies were
receiving both fees for administering their Section 8 contracts.  One
agency, for example, received annual contributions contract fees of
$634,000 and override fees of $584,000 for administering the same
Section 8 contracts during the same period.  The state agencies have
argued, in essence, that they are entitled to both fees because HUD
effectively approved these fees when it approved agreements between
the Department and the agencies to share bond refunding savings
(called McKinney Act refunding agreements).\5 To resolve this issue
for agencies whose refunding agreements it approved, HUD required the
agencies in 1996 to request waivers of its regulation prohibiting
dual fees.  However, as of June 1999, HUD had not taken action on
these requests.  In addition, HUD has not identified all agencies
that are collecting dual fees and has not taken any action when dual
fees are being collected for contracts that are not under refunding
agreements approved by the Department. 

--------------------
\4 Local agencies are not eligible for an override fee because they
issue tax-exempt bonds under the United States Housing Act of 1937
and are subject to HUD's regulations.  State agencies issue
tax-exempt bonds under the Internal Revenue Code. 

\5 Refunding agreements identify the total savings that will become
available from refunding bonds associated with Section 8 contracts
that provide for returning bond refunding savings to the federal
government.  The agreements specify the amounts that will be provided
to the agencies and to the federal government under section 1012 of
the McKinney Act, as amended. 

      HUD'S INFORMATION ON THE
      PHYSICAL AND FINANCIAL
      CONDITION OF THE UNINSURED
      PORTFOLIO IS LIMITED, BUT
      STATE AGENCIES REPORT THAT
      FEW PROPERTIES HAVE PROBLEMS
-------------------------------------------------------- Chapter 0:4.3

While HUD requires annual physical inspections of the properties in
its uninsured portfolio, it did not, until recently, have objective
criteria for ranking the properties' condition.  Thus, although the
ratings in HUD's central database showed that most of the properties
were in satisfactory or better physical condition, the ratings were
subjective and their reliability was therefore limited.  The 10 state
agencies that GAO surveyed reported that 95 percent of the
properties, representing 97 percent of the apartment units, in their
portfolios were in satisfactory or better physical condition. 
Although all 10 agencies used HUD's termsï¿½superior, satisfactory,
below average, or unsatisfactoryï¿½to rate the physical condition of
their properties, their ratings were also subjective.  Consequently,
the state agencies' ratings are subject to the same limitations as
HUD's. 

Information on the financial condition of properties in the uninsured
portfolio is also limited.  Although HUD requires annual audited
financial statements for properties in most of the uninsured
programs, its central database did not, as of December 1998, include
information on the results of these audits.  As a result, overall
conclusions on the financial status of the uninsured portfolio cannot
be drawn at this time.  HUD does not require the state agencies to
assign a rating to the financial condition of their properties. 
Nevertheless, 5 of the 10 state agencies that GAO surveyed had rated
the financial condition of their properties.  These agencies reported
that about 97 percent of their properties were in satisfactory or
better financial condition.  In addition, seven of the agencies had
rated the overall condition of their propertiesï¿½assessing their
management as well as their physical and financial conditionï¿½and
reported that 95 percent were in satisfactory or better overall
condition.  Finally, the 10 state agencies reported that fewer than 4
percent of the properties in their portfolios had problems serious
enough to warrant special monitoring attention. 

In mid-1998, HUD established the Real Estate Assessment Center to
collect and analyze data on multifamily properties in several
portfolios, including the uninsured Section 8 portfolio; develop an
objective system for rating the physical condition of these
properties; and analyze financial information on the properties. 
Currently, trained contractors are inspecting the properties using
the Center's new rating system, and many property owners are required
to submit audited financial statements to the Department
electronically by June 30, 1999. 

   RECOMMENDATIONS
---------------------------------------------------------- Chapter 0:5

This report recommends that the Secretary of Housing and Urban
Development (1) clarify the requirements for state housing finance
agencies to share the savings from refunding bonds with the federal
government, (2) clarify and reissue HUD's guidance on calculating
rent increases when savings have resulted from refunding bonds, and
(3) enforce the Section 8 regulation prohibiting dual fees. 

   AGENCY COMMENTS AND GAO'S
   EVALUATION
---------------------------------------------------------- Chapter 0:6

GAO provided copies of a draft of this report to HUD and to the
National Council of State Housing Agencies\6 for review and comment. 
HUD disagreed with GAO's recommendation that it clarify the
requirements for state housing finance agencies to share the savings
from refunding certain bonds with the federal government.  According
to HUD, the recommendation directs the Department to take action
where legal authority is unclear and proposes that HUD retroactively
recover savings that state agencies have not shared.  However, GAO's
recommendation statedirects HUD to determine whether the state
agencies are required to share certain bond Evaluation). 
tharefunding savings with the government, and if they are, whether
the Department can enforce the requirement prospectively.  However,
GAO recognizes that a sentence in the draft executive summaryï¿½stating
that HUD has not issued guidance to the state agencies directing them
to share certain bond refunding savings--may have implied that the
legal issue had been resolved.  GAO therefore revised this statement
for clarity and greater consistency with the discussion of this issue
in the body of the report (See ch.  3, Agency Comments and Our
Evaluation). 

HUD questioned GAO's support for the statement that, without
clarification of the McKinney Act's shared savings provision, state
agencies may retain ï¿½tens of millionsï¿½ of dollars that they could be
legally required to share with the government.  This figure, cited to
convey the magnitude of the savings at issue, was based on
information on some state agencies' bond refundings, including the
refundings of two agencies that shared savings when they were not
contractually required to do so.  Nonetheless, GAO agrees with HUD
that for contracts that do not include a requirement for providing
bond refunding savings to the government, these savings will
generally be smaller than for contracts that do include this
requirement.  This is because the interest rates--and hence the bond
refunding savings--are generally lower for the contracts without the
requirement.  GAO also agrees with HUD that the data needed to
prepare a comprehensive estimate of the potential savings are not
available.  Therefore, GAO did not include an estimate of potential
savings in the final report and concluded that HUD may have missed
opportunities to provide additional bond refunding savings to the
government. 

HUD agreed with GAO's recommendation on clarifying and reissuing its
guidance on calculating rent increases when savings have resulted
from refunding bonds and plans to implement the recommendation. 
While not disagreeing with GAO's recommendation that the Department
enforce its Section 8 regulation prohibiting dual fees, HUD indicated
that it views the dual fees as an incentive needed for state agencies
to refund their bonds and share the savings with the government. 
However, the agencies were explicitly required by their Section 8
contracts to provide all of the savings to the government in most
instances when shared savings agreements were executed.  In addition,
although the Department says that dual fee transactions will return
savings in excess of $150 million to the Treasury over the life of
the Section 8 contracts, it does not have the information needed to
determine whether these savings will be sufficient to offset the dual
fees provided to the agencies.  GAO did not change its recommendation
in response to these comments. 

Both HUD and the National Council of State Housing Agencies disagreed
with a statement in the draft report that the federal government's
costs are higher than they should be when Section 8 rents exceed
market rents.  GAO revised this statement to describe rather than
evaluate the impact of the Section 8 program's design on the federal
government's subsidy costs.  As revised, the report says that when
Section 8 contract rents exceed market rents, the Section 8 subsidies
support higher rents than the properties generally could command
without federal assistance. 

Like HUD, the National Council disagreed with and misinterpreted
GAO's recommendation that HUD clarify when state agencies are
required to share bond refunding savings with the government. 
According to the National Council, GAO wrongly concluded that HUD has
the authority to require state agencies to share certain bond
refunding savings with the federal government.  In fact, as
discussed, the report recommends that the Secretary determine whether
the state agencies are required to share certain bond refunding
savings.  Nevertheless, as discussed, GAO revised a sentence in the
executive summary that may have caused some confusion.  The National
Council also disagreed with the report for not recognizing that HUD
effectively waived its prohibition of dual fees by approving bond
refunding transactions under which state agencies received both fees. 
As the draft report stated, this is the position of the state
agencies.  However, HUD is required to issue formal waivers when it
does not enforce a regulation such as the prohibition of dual fees. 

HUD's and the National Council's comments and GAO's evaluation of
them are discussed in more detail in chapters 2, 3, and 4 and in
appendixes VI and VI. 

--------------------
\6 The National Council of State Housing Agencies is a national
nonprofit organization that assists state housing agencies in
advancing the interests of lower-income and underserved people
through the financing, development, and preservation of affordable
housing.  Members operate in every state, the District of Columbia,
Puerto Rico, and the U.S.  Virgin Islands. 

BACKGROUND
============================================================ Chapter 1

   INTRODUCTION
---------------------------------------------------------- Chapter 1:1

Section 8 rental housing assistance, managed by the Department of
Housing and Urban Development (HUD), is the main form of federal
housing assistance for low-income tenants.  In fiscal year 1998,
total Section 8 expenditures were about $15.5 billion.  Under Section
8, residents in subsidized units generally pay 30 percent of their
income for rent and HUD pays the balance.  Section 8 rental
assistance is tied either to units in specific properties
(project-based assistance) or to families and individuals who live in
affordable rental housing of their choice (tenant-based assistance). 
Some properties that received project-based assistance also received
federal mortgage insurance through HUD's Federal Housing
Administration (FHA).  The primary goal of the Section 8
project-based program was to encourage developers to build or
rehabilitate properties for lower-income families by providing rental
assistance contracts for a negotiated number of units for periods
ranging from 20 to 40 years.  Authorized in 1974, project-based
assistance was, with one exception, repealed by the Congress in 1983
because of its high cost.  That exception was the assistance used to
provide housing for the elderly and the disabled. 

Over half of HUD's portfolio of approximately 22,000 multifamily
properties that receive Section 8 project-based rental assistance do
not receive federal mortgage insurance.  Collectively, these
properties are referred to as HUD's uninsured Section 8 portfolio.  A
mandate for us to study this portfolio was included in HUD's fiscal
year 1998 appropriations bill (P.L.  105-65, Oct.  27, 1997).  The
uninsured portfolio includes properties for the elderly and disabled
that have been financed by direct loans and capital advances (grants)
from HUD, as well as properties financed by state and local housing
finance agencies, referred to as state and local agencies in this
report. 

      THE UNINSURED SECTION 8
      PROJECT-BASED PORTFOLIO
-------------------------------------------------------- Chapter 1:1.1

Eight programs provide rental assistance to residents of properties
in the uninsured Section 8 project-based portfolio.  As shown in
table 1.1, five of the programs developed housing for low-income
residents using varying financing methods. 

                               Table 1.1
                
                  Housing Development Programs in the
                   Uninsured Section 8 Project-Based
                               Portfolio

Housing development programs                Initial term
with project-based rental     Financing     of Section 8  Program's
assistance                    method        contract      status
----------------------------  ------------  ------------  ------------
Elderly/disabled loan         Government    20 years      Terminated
program                       loans from                  in the early
                              HUD at                      1990s;
                              below-                      replaced
                              market                      with the
                              interest                    capital
                              rates that                  advance
                              were                        program
                              established
                              annually by
                              the Congress

Elderly/disabled capital      Capital       5 or 20       Ongoing
advance program\a             advances      years
                              (grants)      (depending
                              from HUD      on when a
                                            property was
                                            initially
                                            developed)

State agency program          State         20 to 40      Terminated
                              government    years         in 1983
                              tax-exempt
                              bonds

New construction/             Various       20 to 40      Terminated
substantial rehabilitation    methods,      years         in 1983
program                       including
                              local
                              government
                              tax-exempt
                              bonds and
                              some state
                              tax-exempt
                              bonds

Rural rental housing program  Government    20 years      Terminated
                              loans from                  in 1983\b
                              the Rural
                              Housing
                              Service with
                              a 1-percent
                              interest
                              rate
----------------------------------------------------------------------
\a The rental assistance contracts under this program are not funded
under the same appropriations account as the Section 8 rental
assistance program, but the project-based assistance under this
program is substantially the same as Section 8 project-based
assistance except that the subsidy is limited to operating costs. 

\b The Department of Agriculture continues to fund the Rural Rental
Housing development program, but HUD no longer provides new Section 8
project-based assistance.  The Rural Housing Service has its own
rental subsidy program. 

The three remaining programs represented in the uninsured Section 8
project-based portfolio were established in the 1980s to provide
long-term rental assistance to existing properties that were formerly
in FHA's insured portfolio.\1 Section 8 assistance was extended
through these programs as a means of retaining affordable housing for
low-income residents.  First, the loan management set-aside program
provided Section 8 rental assistance to financially troubled projects
using 15-year Section 8 contracts.  Second, the multifamily property
disposition program also used 15-year contracts to provide rental
assistance to properties acquired by new owners through foreclosures
when borrowers defaulted on loans insured by FHA.  Finally, the
housing preservation program provided rental assistance to property
owners who were approaching eligibility to pay off their mortgages as
an incentive for them to maintain some of the units in these
properties as affordable low-income housing.  Preservation contracts
were executed for varying terms, depending on, among other things,
the availability of appropriations.  The Department no longer funds
new project-based contracts under these three programs. 

According to HUD's data, as of December 1998, the uninsured Section 8
project-based portfolio consisted of 12,708 active contracts for
12,488 properties covering 632,216 assisted units.  (See app.  I for
more detailed information, by program, and app.  II for a discussion
of the databases we used to identify the universe of uninsured
Section 8 contracts).\2 Figure 1.1 shows the number and percentage of
assisted units in the uninsured portfolio that are funded through
each of the rental assistance programs.  A majority of the units in
this portfolio--60 percent--are associated with the elderly and
disabled loan program and the state agency program. 

   Figure1.1:  Percentages and
   Numbers of Assisted
   Project-Based Units, by Program

   (See figure in printed
   edition.)

Note:  Percentages do not add because of rounding. 

Source:  HUD's Real Estate Management System and Section 8 expiring
contracts databases, as of Dec.  1998. 

--------------------
\1 The uninsured portfolio also includes properties that receive
rental assistance under HUD's rental assistance payment and rent
supplement programs (representing 371 contracts covering 30,250
units).  These programs, which were precursors of the Section 8
program, are not discussed in this report. 

\2 This universe includes 13,046 active contracts, 12,708 of which
cover properties in the eight programs discussed in this report. 

      BOND FINANCING FOR
      PROPERTIES IN THE UNINSURED
      PORTFOLIO
-------------------------------------------------------- Chapter 1:1.2

Some properties in the uninsured portfolio were financed by state and
local housing finance agencies with the proceeds of bonds that are
exempt from federal taxation.  Specifically, during the 1970s and
early 1980s, tax-exempt bonds were used to finance the development of
the 2,278 properties in the state agency program and a portion of the
1,678 properties in the new construction and substantial
rehabilitation program.\3 State housing finance agencies (state
agencies) generally issued bonds as instrumentalities of the state
under section 103 of the Internal Revenue Code.  Section 8 assistance
for properties financed by state agencies was generally approved from
Section 8 funds allocated to these agencies and is identified in
HUD's data systems under the state agency program.  Local agencies
and instrumentalities (local agencies) generally issued tax-exempt
bonds under Section 11(b) of the United States Housing Act of 1937. 
The state and local agencies used the bond proceeds to provide
mortgages for constructing or substantially rehabilitating properties
for use as affordable housing for low-income people under the Section
8 program.  The monthly mortgage payments are used to repay the bonds
with interest. 

During the late 1970s and early 1980s, interest rates rose to
unprecedented levels.  To continue the development of affordable
rental housing despite rising interest costs, HUD authorized special
financing in 1980 and again in 1981.  This special financing allowed
for higher mortgage interest rates, which then increased the costs of
HUD's Section 8 rental assistance.  The Section 8 contracts that
received the special financing authorized in 1980 did not require
state and local agencies to refund (refinance) their high-interest
bonds if interest rates later dropped.  But as interest rates
remained high and HUD concluded that special financing would be
required for an extended period of time, the Department took steps to
reduce Section 8 costs in the future when interest rates declined. 
As a result, the Section 8 contracts that received special financing
beginning in October 1981ï¿½called financing adjustment factor
contracts--did require state and local agencies to refund their bonds
when interest rates fell and provide the savings to the government. 
HUD also required property owners with financing adjustment factor
contracts to accept reduced Section 8 contract rents (subsidies) to
reflect the decrease in borrowing costs resulting from refunding the
bonds. 

In 1987, after interest rates started to decline, HUD asked the
housing agencies to refund the bonds associated with financing
adjustment factor contracts.  The savings from bond refundings can be
substantial.  For example, a local agency refunded bonds for three
mortgage loans totaling over $14 million.  The savings from these
refundings, which will be realized over the lives of the mortgages,
are estimated to be $6.4 million.  These savings represent the
difference between the cost of the mortgages needed to repay the
original bonds and the cost of the mortgages needed to repay the
refunded bonds. 

The Congress, in passing the Stewart B.  McKinney Homeless Assistance
Amendments Act of 1988, approved a provision that permitted state
agencies to keep half of the savings from refunding bonds associated
with financing adjustment factor contracts.  While the McKinney Act
provided primarily for assistance to the nation's homeless
population, section 1012 created an incentive for the state agencies
to refund bonds associated with these contracts.  Without the
McKinney Act's shared savings provision, the state agencies would
have been contractually required to provide 100 percent of the
savings to the government. 

As amended in April 1992, section 1012 extended this benefit to local
agencies.  Finally, as amended in October 1992, the section provided
for state and local agencies and the federal government to share the
savings from refunding bonds associated with Section 8 contracts
entered into between 1979 and 1984.  This period generally covered
the contracts that received special financing, including the
financing adjustment factor contracts, and some contracts that did
not receive special financing. 

In 1992 and 1993, HUD's Office of the Inspector General (OIG)
released two reports that examined whether (1) bond-financed Section
8 properties were refinanced as intended and (2) HUD realized the
appropriate savings from the bond refundings.\4 The initial report,
which included 18 recommendations, was issued as an interim report
before the audit work was completed because the Inspector believed
the potential for cost savings and the need to improve internal
controls warranted prompt corrective action.  The final report,
issued on April 30, 1993, included six additional recommendations. 
The reports disclosed, among other things, that HUD had not fully
realized potential savings from bond refundings and identified
actions that HUD could take to realize additional savings. 

In January 1997, the Inspector General determined that HUD had not
taken the corrective actions it had agreed to take in response to
several of the key recommendations in these reports.  Therefore, the
Inspector General reopened the recommendations.  The reopened
recommendations include those addressing the extent to which agencies
are required to share bond refunding savings with the government,
excess rent increases to owners, and dual fees for administrating
Section 8 contracts.  As of June 1999, these issues had not been
effectively resolved.  These issues are discussed further in chapter
3. 

--------------------
\3 While the Section 8 contracts associated with properties financed
by local agencies are included with others under the overall new
construction and substantial rehabilitation program, HUD's Section 8
data do not identify those with bonds issued by local agencies.  This
program also includes some properties financed by state agencies that
were not processed under the state agency program. 

\4 Interim Audit Report Bond Refundings of Section 8 Projects (OIG
93-HQ-119-0004, Oct.  30, 1992) and Multi-Region Audit of Refunding
of Bonds For Section 8 Assisted Projects (OIG 93-HQ-119-0013, Apr. 
30, 1993). 

      HOW BOND REFUNDING SAVINGS
      ARE SHARED
-------------------------------------------------------- Chapter 1:1.3

To share the savings from refunding bonds, HUD enters into agreements
with state and local agencies that identify (1) the total amount of
savings that will become available as a result of the refundings and
(2) the amounts that will be provided to the agencies and to the
federal government each year throughout the lives of the Section 8
contracts.  HUD refers to the shared savings agreements as refunding
agreements.  According to HUD's data, as of September 30, 1998, the
Department had approved 245 refunding agreements that will provide
$1.1 billion in bond refunding savings over the lives of the Section
8 contracts, $633 million of which is to be provided to the U.S. 
Treasury.  In negotiating refunding agreements with agencies, HUD
reviews the agencies' bond refunding documents.\5

The responsibility for making the shared savings payments varies,
depending on the method selected to share the savings.  Two methods
are available for state and local agencies and the federal government
to share savings.  The first, called the rent reduction method,
reduces the federal government's Section 8 costs directly.  Under
this method, the mortgage is refinanced and the Section 8 contract
rents are reduced to reflect the new, lower cost of bond financing. 
HUD periodically pays the agency its share of the savings that accrue
over the life of the mortgage and the Section 8 contract.  A few
agencies have used this method to share savings. 

An alternative method of sharing savings, called the trustee sweep
method, is used by most state and local agencies that share savings,
according to HUD.  Under this method, neither the mortgage nor the
Section 8 contract rents are reduced after the bond is refunded. 
Instead, an independent third partyï¿½a trusteeï¿½receives the mortgage
payments from the Section 8 property owners and uses these funds to
repay the bond principal and interest.  The remaining balance (bond
payments minus mortgage payments) represents the savings from
refunding the bonds.  Semiannually, the trustee pays the state or
local agency its share of the savings and pays the federal share to
the U.S.  Treasury.  Thus, the federal government receives
reimbursement for a portion of the excess financing costs being borne
by the Section 8 program. 

Using the trustee sweep method eliminates the need for agencies to
refinance mortgages and amend their Section 8 contracts with owners
and for HUD to set up accounts to pay the state or local agencies
their share of the bond refunding savings.  Thus, it imposes less of
an administrative burden on both the state agencies and the
Department.  According to HUD officials, many agencies would not have
participated in the refunding program if the trustee sweep method had
not been available.  However, the trustee sweep method allows Section
8 costs (expenditures) to remain artificially highï¿½that is, the costs
continue to reflect the original high interest rates.  As is
discussed further in chapter 3, the trustee sweep method can result
in excess rent increases for many contracts.  This occurs because
many Section 8 contracts receive automatic rent increases on the
basis of a factor that is applied to total Section 8 costs-that is,
to the debt service as well as the operating costs. 

--------------------
\5 State agencies issuing bonds under IRS' regulations did not
usually need HUD's approval to issue bonds.  As a result, the
Department generally does not review state agency bond transactions. 
However, HUD does review the state agency bond refunding transactions
associated with McKinney Act refunding agreements. 

      CONTRACT ADMINISTRATION FEES
      IN THE UNINSURED PORTFOLIO
-------------------------------------------------------- Chapter 1:1.4

State and local housing finance agencies that administer Section 8
contracts for HUD receive compensation for carrying out their
administrative responsibilities.  This compensation is paid to the
agencies for performing administrative tasks, such as conducting
management reviews of the Section 8 properties and inspecting the
properties at least annually.  All of the agencies are eligible for
what is called an annual contributions contract fee.  Alternatively,
state agencies that finance Section 8 property mortgages with the
proceeds of state tax-exempt bonds issued under the Internal Revenue
Service's (IRS) regulations may receive what is referred to as an
ï¿½override feeï¿½ instead of an annual contributions contract fee. 

The annual contributions contract fee is generally a per-unit cost
(equal to 3 percent of the annual fair market rent for a 2-bedroom
unit) multiplied by the number of units in the property.  An override
fee represents the difference between the agency's borrowing (bond
issuance) and mortgage lending rates.  This difference, which IRS
refers to as arbitrage, cannot exceed 1.5 percent for all of the
properties the agency has financed through the bond proceeds. 
Because local housing finance agencies generally issued bonds under
the United States Housing Act of 1937, they are not eligible for the
override fee available to state housing finance agencies under IRS'
regulations.  Under a Section 8 regulation promulgated in 1980 (24
C.F.R.  883.606), a state housing finance agency that chooses to
collect an override fee cannot receive an annual contributions
contract fee. 

      OBJECTIVES, SCOPE, AND
      METHODOLOGY
-------------------------------------------------------- Chapter 1:1.5

Section 532 of the 1998 Appropriations Act for the Departments of
Veterans Affairs and Housing and Urban Development, and Independent
Agencies (P.L.105-65, Oct.  27,1997) requires that GAO submit a
report to the Congress on the uninsured Section 8 portfolio. 
Accordingly, this report examines (1) the information HUD has on the
Section 8 assistance provided to properties in the uninsured
portfolio; (2) the financial benefits that may be available to state
and local housing finance agencies that participate in the Section 8
program and the impact of these benefits on the Section 8 program's
costs; and (3) the ways HUD and the state agencies oversee the
physical and financial condition of the properties in their
respective uninsured Section 8 portfolios and the information HUD and
state agencies have on the physical and financial condition of these
properties. 

To determine what information the Department maintains on the Section
8 assistance provided to uninsured properties and the physical and
financial condition of these properties, we obtained several HUD
databases that were used to develop the information responding to
these objectives and met with officials from HUD's Office of Housing
and Office of Chief Financial Officer.  (See app.  II for additional
information on the databases used in this review.) We also obtained
information from HUD officials on the improvements the Department is
implementing for monitoring the physical and financial condition of
its properties under the Real Estate Assessment Center.  We did not
evaluate the effectiveness of HUD's new processes. 

To develop information on the benefits available to state and local
agencies that participate in the uninsured Section 8 program and the
impact of these benefits on the Section 8 program's costs, we
interviewed officials from HUD, state agencies, and the National
Council of State Housing Agencies.  We sent a data collection
instrument to state housing finance agencies in Illinois, Maryland,
Massachusetts, Minnesota, and New Hampshire.  We also reviewed HUD
documents--including legal opinions, regulations, notices, and
handbooks--and literature on tax-exempt bonds from bond-rating
agencies. 

To identify the ways state agencies monitored uninsured project-based
Section 8 portfolios and ascertained the physical and financial
condition of their properties, we selected five state agencies in
addition to the five that provided information on Section 8
benefitsï¿½California, Michigan, Oregon, Tennessee, and Wisconsin. 
These 10 state agencies have portfolios of varying sizes and serve
different geographical regions, including rural and urban areas.  We
sent a data collection instrument to the 10 state agencies to obtain
information on (1) their overall monitoring approaches and the
primary methods they use to evaluate the physical and financial
condition of their portfolios and (2) the physical and financial
condition of the uninsured Section 8 properties in their portfolios. 
In addition, we visited the Illinois, Maryland, Massachusetts,
Minnesota, and New Hampshire agencies, where we discussed monitoring
approaches, reviewed specific project files, and obtained
documentation on policies and procedures. 

We conducted our work from August 1998 through July 1999 in
accordance with generally accepted government auditing standards. 

   AGENCY COMMENTS
---------------------------------------------------------- Chapter 1:2

We provided copies of a draft of this report to HUD and to the
National Council for State Housing Agencies for review and comment. 
Their comments, which are reproduced in appendixes VI and VII, are
discussed and evaluated as applicable in the remaining chapters of
this report. 

SECTION 8 ASSISTANCE PROVIDED TO
PROPERTIES IN THE UNINSURED
PORTFOLIO
============================================================ Chapter 2

According to HUD's data, rental assistance payments for the uninsured
Section 8 portfolio totaled over $3.3 billion in fiscal year 1998.  A
majority of these payments-about $2.3 billion--were associated with
the two largest programs in the uninsured portfolio, the state agency
and the elderly/disabled loan programs.  The average payment per
rental unit varied significantly from program to program, reflecting
in large measure the different financing methods used in the various
programs.  Complete data were not available for comparing assisted
rents to market rents-the commonly accepted standard for assessing
the reasonableness of rents.  Nevertheless, on the basis of
information that is available, some assisted rents exceed market
rents.  When Section 8 contract rents exceed market rents, the
Section 8 subsidies support higher rents than the properties
generally could command without federal assistance.  Moreover, the
federal government will continue to incur these high rent costs each
year until its existing Section 8 contracts expire.  These contracts
will expire at various times, from within the next 5 years to about
20 years.  Contracts in the state agency program will generally be
among the last to expire. 

   SECTION 8 EXPENDITURES AND
   PER-UNIT COSTS VARY WIDELY BY
   PROGRAM
---------------------------------------------------------- Chapter 2:1

Section 8 rental assistance payments for the uninsured portfolio
totaled over $3.3 billion during fiscal year 1998.  This amount
represents net expenditures--that is, fiscal year 1998 expenditures
to property owners minus any offsetting collections received during
the period.  As shown in figure 2.1, about 68 percent of the
portfolio's rental assistance expenditures are distributed among the
two largest programs in the uninsured portfolio--the state agency
program and the elderly/disabled loan program. 

   Figure 2.1:  Section 8 Contract
   Expenditures for Fiscal Year
   1998, by Program

   (See figure in printed
   edition.)

Note:  Percentages do not add because of rounding. 

Source:  Extract of fiscal year 1998 expenditures and receipts from
HUD's program accounting system. 

The average per-unit subsidy costs for fiscal year 1998 for the eight
programs also varied considerably.  As shown in table 2.1, the
average per-unit costs ranged from $1,492 for the elderly/disabled
capital advance program to $6,903 for the state agency program. 

                               Table 2.1
                
                   Average Per-Unit Subsidy Cost, by
                  Section 8 Program, Fiscal Year 1998

Program                                          Average per-unit cost
----------------------------------------  ----------------------------
Elderly/disabled capital advance                                $1,492
Loan management set-aside                                        2,966
Rural housing                                                    4,281
Housing preservation                                             4,696
Multifamily property disposition                                 4,870
Elderly/disabled loan                                            5,374
New construction/substantial                                     5,421
 rehabilitation
State agency                                                    $6,903
----------------------------------------------------------------------
Source:  Extract of fiscal year 1998 expenditures and receipts from
HUD's program accounting system. 

An important reason for the variation in average per-unit costs is
that the programs were financed in different ways.  For example, the
federal government provides grants to develop properties under the
elderly/disabled capital advance program.  Because the government
makes an investment up front, the Section 8 program's subsidies need
to cover only the operating costs for properties in this program.  In
contrast, properties under the other development programs were
financed with mortgage loans.  Their Section 8 subsidies are
considerably higher because they must cover both the mortgage debt
service and the operating costs. 

Another factor that affects the level of Section 8 subsidies is
tenants' income.  As discussed in chapter 1, residents generally pay
30 percent of their income for rent, and HUD pays the balance.  As a
result, differences in tenants' income levels can influence the
average per-unit costs for these programs.  Thus, one factor
contributing to the high per-unit costs in the state agency program
may be the residents' low income levels.  According to the 10 state
agencies we surveyed, about 93 percent of the households receiving
Section 8 project-based rental assistance at their multifamily
properties had very low incomes--defined by HUD as at or below 50
percent of the local area's median income. 

Officials from HUD's Office of Housing, including the Directors for
Business Products and Portfolio Management, identified these and
other financing and programmatic differences, summarized below, that
can affect the per-unit cost of the programs. 

  -- The state agency program allowed more flexibility than most of
     the other programs in setting initial contract rents.  The rents
     could be higher than those permitted under HUD's standard new
     construction/substantial rehabilitation Section 8 program. 
     According to Office of Housing officials, the higher rent
     structure allowed the state agencies to develop properties that
     offered more amenities than the typical Section 8 property
     insured by FHA.  In particular, the officials said, the higher
     rents allowed the state agencies to develop some properties in
     affluent suburban neighborhoods that were compatible with the
     housing in those neighborhoods.  The officials also said that a
     number of state agency properties received a financing
     adjustment factor that allowed higher contract rents to support
     the high mortgage interest rates prevailing during the early
     1980s. 

  -- The elderly/disabled loan program also allowed greater
     flexibility in setting initial contract rents, which often
     exceeded market rents.  In addition, the properties were
     generally more costly to develop because they were mid-rise
     and/or high-rise buildings that provided more amenities, such as
     emergency call systems, than most of the other subsidized new
     construction/substantial rehabilitation properties.  These
     higher costs are offset to some extent by below-market interest
     rates, which the Congress established annually for the program. 
     Although interest rates were below market when the financing for
     the properties was approved, some of the rates are now higher
     than current market interest rates. 

  -- The rural housing program serves low-income persons, including
     the elderly and disabled, but properties under this program have
     lower per-unit costs than similar properties in HUD's uninsured
     portfolio.  According to HUD officials, the rural housing
     properties are older than the HUD properties and are built in
     rural areas, where construction costs are generally lower than
     in urban areas.  In addition, the properties were financed with
     loans that generally had subsidized interest rates of 1 percent. 

Each program's per-unit costs reflect the influence of a variety of
factors on the long-term costs of Section 8 rental assistance, and
each program's costs need to be evaluated in the context of these
factors.  Furthermore, the per-unit costs may not reflect all of the
costs that the government incurs for some programs.  For example, the
per-unit costs for grant programs, such as the elderly/disabled
capital advance program, and for interest subsidy programs, such as
the rural housing program, do not include the costs of the federal
grants or interest subsidies provided under these programs. 
Additional data--which may or may not be available for all of the
programs--and in-depth analyses would be required to estimate the
total per-unit costs of these programs to the government. 

   HUD'S RENT STANDARD IS USED TO
   LIMIT RENT INCREASES, BUT
   PORTFOLIO DATA ON ACTUAL MARKET
   RENTS ARE NOT AVAILABLE
---------------------------------------------------------- Chapter 2:2

The Congress, beginning in fiscal year 1995, limited the annual rent
increases that, until then, were automatic for many Section 8
properties.  Under this congressional limit, automatic rent increases
are no longer allowed for properties whose Section 8 contract rents
exceed HUD's rent standard--fair market rents--unless the property
owners provide independent studies showing that the Section 8
contract rents do not exceed actual market rents.\1 To establish fair
market rents, HUD annually samples market rents for geographic areas,
such as large metropolitan areas, and sets the fair market rent
somewhat below the average for the geographic area.  In some cases,
the area covered by HUD's fair market rent is too wide to reflect
differences in the rents paid in different submarkets, or
neighborhoods, within the geographic area covered.  By contrast,
market rents reflect the rents paid for comparable units in
particular neighborhoods.  As a result, HUD's fair market rents for
Section 8 properties may not reflect the actual market rents in
neighborhoods where Section 8 properties are located. 

According to HUD's data, most of the rents for Section 8 units in the
uninsured portfolio exceed HUD's fair market rents and are therefore
subject to the congressional limit on rent increases.  As of December
1998, the rents for 474,270 of 632,216 assisted units in the
uninsured portfolio exceeded fair market rents.  These units
represent 75 percent of the assisted units in the uninsured
portfolio.  As shown in figure 2.2, for 22 percent of these 474,270
units, the Section 8 rents were greater than 160 percent of HUD's
fair market rents. 

   Figure 2.2:  Unit Rents
   Exceeding HUD's Fair Market
   Rent Levels, December 1998

   (See figure in printed
   edition.)

FMR=fair market rent

Source:  HUD's Section 8 expiring contracts database, Dec.  1998. 

Among the uninsured programs, those that use Section 8 assistance to
support property mortgages generally have rents that exceed fair
market rents.  Thus, most of the units in the four housing
development programs -- the state agency, elderly/disabled loan,
rural housing, and new construction/substantial rehabilitation
programsï¿½have Section 8 contract rents that exceed HUD's fair market
rents.  These programs account for over 80 percent of the assisted
units in the entire uninsured inventory.  As shown in table 2.2, the
rents for between 72 and 91 percent of the units associated with the
four programs exceeded HUD's fair market rents.  (See app.  III,
table III.1 for additional details on these rent levels by program.)

                               Table 2.2
                
                   Percentage of Assisted Units With
                  Section 8 Rents Exceeding HUD's Fair
                   Market Rents, by Program, in HUD's
                Uninsured Section 8 Portfolio, December
                                  1998

                                          Percentage of units with
                                          rents exceeding fair market
Program                                   rents
----------------------------------------  ----------------------------
State agency                              91

Elderly/disabled loan                     88

Rural housing                             86

New construction/substantial              72
rehabilitation

Preservation                              42

Loan management set-aside                 24

Property disposition                      21

Elderly/disabled capital advance          5
----------------------------------------------------------------------
Note:  These data cover 12,288 of the 12,708 contracts in HUD's
uninsured portfolio for which data on rents and HUD's fair market
rents were available. 

Source:  HUD's Section 8 expiring contracts database, Dec.  1998. 

Because data on actual market rents for particular neighborhoods were
not available for comparison with Section 8 contract rents, we could
not determine to what extent the contract rents exceeded the actual
market rents in this portfolio.  However, some properties in the
uninsured Section 8 portfolio have rents in excess of actual market
rents. 

When Section 8 contract rents exceed market rents, the Section 8
subsidies support higher rents than the properties generally could
command without federal assistance.  In the uninsured portfolio, as
in the FHA-insured portfolio, where HUD found that many of its
Section 8 rents exceeded market rents,\2

the assisted rents were initially set above market levels to
encourage the production of new affordable housing.  These rents were
then increased automatically each year through the application of set
formulas that, according to HUD, tended to be generous.\3 A July 1997
study of 53 bond-financed properties, the majority of which are
included in the uninsured Section 8 portfolio, reported that some of
the properties had rents that were well above comparable market
levels.\4

In addition, since the Congress limited automatic Section 8 rent
increases, some owners have not requested rent increases.  Officials
at two of the five state agencies we visited said they have not
processed Section 8 rent increases for owners because the Section 8
rents for properties in their portfolios are generally higher than
market rents.  Officials at another state agency we visited said that
before the Congress limited automatic rent increases, they had
attempted to increase rents only for the uninsured Section 8
properties in their portfolio that needed increases to cover their
costs.  However, as long as the Section 8 program's rules required
automatic annual rent increases, the agency had to provide them
according to a formula that sometimes provided more assistance than
the officials considered necessary. 

Besides limiting automatic rent increases, the Congress has taken
some other steps to reduce high Section 8 costs.  For example, the
Congress enacted ï¿½mark to market,ï¿½ or multifamily portfolio
reengineering, legislation in 1997 to bring rents in the FHA-insured
Section 8 portfolio in line with market rents.  Under this
legislation, Section 8 contract rents are to be reset to market
levels and mortgage debt is to be reduced if necessary to permit a
positive cash flow.  These efforts were designed not only to reduce
the costs of expiring Section 8 contracts but also to address
problems at financially and physically troubled projects, correct
management and ownership deficiencies, and preserve the affordability
and availability of low-income rental housing.  More recently, in
HUD's fiscal year 1998 appropriations legislation, the Congress
placed limits on the contract rents that will be allowed when Section
8 contracts in the uninsured Section 8 portfolio expire and are
renewed.  According to the legislation, contract renewal rents are
authorized at the lower of (1) a level that provides sufficient
income to support rents based on actual costs (budget-based rents) or
(2) existing rents subject to a new adjustment factor that allows
increases on operating costs, in contrast to the current factor that
allows increases on both operating and debt service costs. 

--------------------
\1 In the tenant-based section 8 program, fair market rents are used
as limits for rents the Department will subsidize. 

\2 Our 1996 report on HUD's portfolio reengineering proposal
discusses high rents in the FHA-insured portfolio, and a study by
Ernst & Young for HUD on the potential cost to the government of
reducing Section 8 rents in the insured portfolio to market rents. 
See Multifamily Housing:  Effects of HUD's Portfolio Reengineering
Proposal, GAO/RCED-97-7, Nov.  1996). 

\3 A problem with automatic rent increases, according to Office of
Housing officials, is that they allow a percentage increase in both
operating costs and mortgage debt service, even though debt service
is a fixed amount that does not increase. 

\4 HUD's Local Multifamily Portfolio, John Nuveen & Co., Inc.  (July
1997). 

   SECTION 8 CONTRACTS FOR
   PROGRAMS IN THE UNINSURED
   PORTFOLIO WILL EXPIRE AT
   VARIOUS TIMES
---------------------------------------------------------- Chapter 2:3

As discussed in chapter 1, the Section 8 contracts for properties in
the uninsured portfolio were initially made for periods ranging from
20 to 40 years.  Currently, the contracts covering 42 percent of the
assisted units will expire by the end of 2004.  However, a number of
contracts in the uninsured portfolio have 15 to 20 years remaining
(see fig.  2.3).  A majority of the assisted units--about 58
percent--are under contracts that will expire between 2005 and 2038. 
As a result, the federal government will continue to incur costs
under these rental assistance contracts for many years. 

   Figure 2.3:  Dates When Section
   8 Contracts for Uninsured
   Properties Will Expire

   (See figure in printed
   edition.)

Note:  Excluded from this analysis are 1,795 units covered by 54
Section 8 contracts whose expiration dates are unknown. 

Percentages do not add because of rounding. 

Source:  GAO's analysis of HUD's Section 8 expiring contracts
database, Dec.  1998. 

From program to program, contracts will expire at different times. 
For example, the vast majority of the contracts for the loan
management set-aside program and the rural housing program will
expire between 1999 and 2004, while most of the contracts for the
state agency program will expire laterï¿½-between 2010 and 2021.  (See
app.  III, table III.2, for additional information on when contracts
will expire, by program.)

   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 2:4

In commenting on the draft report, HUD and the National Council
disagreed with our statement that the federal government's costs are
higher than they should be when Section 8 rents exceed market rents. 
We revised this statement to describe rather than evaluate the impact
of the Section 8 program's design on the federal government's subsidy
costs.  As the draft report observed, HUD originally set Section 8
rents above market rents to encourage the production of low-income
housing and then automatically increased the rents each year until
1995, using formulas that tended to be generous.  As a result,
Section 8 rents sometimes exceed market rents.  While the draft
report concluded that, in these instances, the program's costs are
higher than they should be, the revised report observes that when
Section 8 rents exceed market rents, the Section 8 subsidies support
higher rents than the properties generally could command without
federal assistance. 

HUD further disagreed with our comparison of Section 8 rents with
HUD's fair market rents.  The draft report recognizes that HUD's fair
market rents are not equal to market rents and describes the
methodology used to set the fair market rent for a particular area at
a level below the average rent for that area.  Moreover, as the draft
report states, national data on market rents were not available. 
Despite their limitations, fair market rent data are the only
comparative rent data that HUD maintains for all Section 8 contracts. 
The Department itself compares fair market rents with Section 8
rents, posting the results in a database on its Internet Web site. 
In addition, the Congress requires that fair market rents be used to
limit some Section 8 rent increases.  Therefore, we did not revise
the report in response to this comment. 

Finally, according to HUD, the draft report provided no evidence,
apart from the fair market rent data, that some Section 8 rents were
higher than market rents.  However, in addition to noting the impact
of the program's design on subsidy and rent levels, the draft report
cited (1) a July 1997 study of 53 bond-financed Section 8 properties
that found the rents for some were well above comparable market
levels, (2) officials at two state agencies who said that the Section
8 rents for properties in their portfolios were generally higher than
market rents, and (3) an official at another state agency who said
that HUD's former rent increase formula sometimes required the agency
to provide more rental assistance than the officials considered
necessary.  In light of this evidence, we did not revise the report
in response to HUD's comment. 

HUD HAS MISSED OPPORTUNITIES TO
REDUCE ITS COSTS IN THE STATE
AGENCY PROGRAM
============================================================ Chapter 3

Financial benefits available to state and local agencies under the
Section 8 program include savings from refunding bonds and a fee for
administering Section 8 contracts.  Agencies use these benefits to
provide affordable housing for low-income residents in their states. 
We found that HUD has not resolved three long-standing issues
associated with these financial benefits, missing opportunities to
reduce its Section 8 costs.  First, HUD has not provided the state
agencies with a clear interpretation of an amendment to the McKinney
Act that provides for agencies to share savings with the government
from refunding bonds associated with certain Section 8 contracts.  As
a result, after refunding bonds, some state agencies have retained
all of the savings that accrue annually, while other agencies have
shared the savings with the government.  Second, HUD has not provided
the agencies with clear guidance for calculating rent increases after
bonds have been refunded.  Without such guidance, some agencies have
not reduced the costs of financing in their calculations and have
thus approved unduly high rent increases.  Finally, some state
agencies have been collecting two fees for administering their
Section 8 contracts, despite a long-standing HUD regulation
prohibiting dual fees.  Because HUD has not enforced its regulation,
the federal government has provided excess funding to state agencies
for administering Section 8 contracts.  Taking action on each of
these issues could reduce the Section 8 costs borne by the federal
government. 

   REQUIREMENTS FOR SHARING
   REFUNDING SAVINGS REMAIN
   UNCLEAR
---------------------------------------------------------- Chapter 3:1

After more than 6 years, HUD has not clarified a provision for state
agencies to share certain bond refunding savings with the government. 
This provision is set forth in section 1012 of the McKinney Act, as
amended in October 1992.  On its face, the amendment appears to
require the state agencies to share savings; however, the National
Council of State Housing Agencies and some state agencies maintain
that the amendment applies only to Section 8 contracts that
explicitly give HUD the right to bond refunding savingsï¿½that is, only
to financing adjustment factor contracts.  While a HUD legal opinion
stated that, on the basis of express language of the statute, the
Department could apply the amendment to state agencies for all
contracts issued between 1979 and 1984, the Department has not issued
clear guidance to this effect.  Over the years, many state agencies
have refunded bonds, often retaining all of the savings.  Meanwhile,
HUD's inaction has potentially deprived the federal government of
opportunities to share savings and reduce Section 8 costs. 

      SECTION 1012 OF THE MCKINNEY
      ACT PROVIDES FOR SHARING
      BOND REFUNDING SAVINGS
-------------------------------------------------------- Chapter 3:1.1

Section 1012 of the Stewart B.  McKinney Homeless Assistance
Amendments Act of 1988 permitted state agencies to keep half of the
savings from refunding bonds associated with financing adjustment
factor contracts to provide affordable housing for households with
very low incomes.  Financing adjustment factor contracts required
agencies to refund high-interest bonds when interest rates declined
and to provide all of the interest savings to the federal government
(see ch.  1).  Without the McKinney Act's shared savings provision,
the state agencies would have been required to provide 100 percent of
the savings to the government under these contracts with HUD.  The
contracts also required property owners to accept reduced Section 8
contract rents (subsidies) to reflect the decrease in borrowing costs
resulting from refunding the bonds. 

In 1992, the Congress amended section 1012 twice.  First, in April
1992, the Congress expanded the shared savings provision for
financing adjustment factor contracts to local agencies.  Then, in
October 1992, the Congress authorized both state and local agencies
to share the savings from refunding bonds associated with Section 8
contracts entered into between calendar years 1979 and 1984.  The
Section 8 contracts executed during this 6-year period include (1)
the financing adjustment factor contracts, (2) the contracts that
received special financing authorized in 1980 when interest rates
were high but do not include a contractual requirement to provide any
bond refunding savings to the government, and (3) some contracts that
did not receive special financing. 

Both of the 1992 amendments to section 1012 provided the local
agencies, which generally issued bonds under HUD's regulations, with
a new financial benefit.  Before the amendments were passed, HUD had
required these agencies to return all bond refunding savings to the
government--whether their Section 8 contracts required them to do so
or not.  After the amendments were passed, the local agencies were
entitled to half, rather than none, of the savings.  However, section
1012, as amended in October 1992, if applied to state agencies, would
reduce by half the financial benefit that some state agencies are
generally accustomed to receiving for refunding bonds associated with
all Section 8 contracts except financing adjustment factor contracts. 
These state agencies generally retain all of the savings except for
financing adjust factor contracts.\1 HUD officials said the
Department does not have the leverage over state agencies that it has
over local agencies because it does not have the authority to review
and approve most state bond issuances.  HUD officials also said that,
in general, the state agencies receive wide latitude and
responsibilities under the Section 8 program while HUD's role is
confined to receiving certifications that the program has been
executed and, occasionally, to auditing the agencies' activities. 
HUD does, however, review information on state agencies' bonds before
entering into agreements (called refunding agreements) with the
agencies to share bond refunding savings with them under the McKinney
Act. 

--------------------
\1 To preserve developments--avoid loan prepayments and conversions
of Section 8 properties to market-rate housing--one of the five state
agencies in our study (Illinois) provided savings to some property
owners in the form of loans.  These loans will be paid over time from
Section 8 rental assistance payments. 

      HUD NEVER RESOLVED
      UNCERTAINTY ABOUT
      CONGRESSIONAL INTENT
-------------------------------------------------------- Chapter 3:1.2

We found that although the Department has expressly required the
state agencies to comply with the shared savings requirements in
section 1012 covering financing adjustment factor contracts, it has
not expressly required the state agencies to comply with the shared
savings provision in the October 1992 amendment covering Section 8
contracts executed between 1979 and 1984.  HUD officials told us that
many state agencies have consistently questioned HUD's authority to
require the state agencies to share savings from refundings that do
not require HUD's approval.  HUD officials also said that the scant
legislative history of the October 1992 amendment focuses on
providing a benefit to local agencies and says nothing about the
state agencies, raising a question as to whether the Congress
intended the amendment to affect the savings from refundings that
state agencies carry out under IRS' regulations--refundings that do
not require HUD's review and approval. 

In May 1993, in response to a request from the Office of Housing,
HUD's Office of the General Counsel issued a formal legal opinion
addressing this question.  The legal opinion concluded that HUD could
require the state agencies to share savings from refundings
associated with all contracts covered by the October 1992 amendment. 
However, the May 1993 legal opinion acknowledged that state housing
finance agencies might challenge HUD if the Department sought to
enforce the clear facial reading of the statute because the
legislative history did not address state agencies.  The Office of
Housing was concerned about the Department's ability to prevail in a
lawsuit over the issue.  From October 1992 until early 1996, the
Department took no steps to require the state agencies to share the
savings from refunding bonds associated with contracts that (1)
received the special financing authorized in 1980 but do not include
a contractual requirement to provide the bond refunding savings to
the government or (2) were entered into between 1979 and 1984 but did
not receive special financing. 

In April 1996, HUD published regulations that were intended to
clarify the rules that applied to state and local agencies.  However,
from the start, the rulemaking caused confusion for the state
agencies because HUD added language to a section of the Code of
Federal Regulations that otherwise applies primarily to local
agencies issuing bonds under the United States Housing Act of 1937. 
The Office of Housing official in charge of the rulemaking said that
the new rule was intended to require state agencies to begin sharing
bond refunding savings in compliance with the October 1992 amendment. 
However, he and other HUD officials agreed that the regulations did
not meet their stated objective because of a typographical error. 
Instead of citing two relevant paragraphs of the McKinney Act, the
published regulations cite two paragraphs that do not apply to state
agencies.  The two relevant paragraphs that HUD intended to cite (1)
identify the universe of contracts subject to the shared savings
requirements as those issued between 1979 and 1984 and (2) specify
the purposes for which McKinney Act savings can be used.  The Office
of Housing official in charge of bond refundings said that no state
agencies have initiated shared savings agreements for new bond
refundings since this regulation was passed. 

In addition, HUD officials acknowledged that by (1) amending the
rules that generally applied only to local agencies and (2) not
clearly stating in the rulemaking that the Department was extending
the requirements for sharing savings to all state agency contracts
covered by the October 1992 amendment to the McKinney Act, HUD might
not have resolved the confusion and controversy over this issue even
if it had not made a typographical error in the final rule.  The HUD
officials were uncertain how to correct the rulemaking.  HUD
officials in the Office of the General Counsel and the Office of
Housing responsible for interpreting and implementing the McKinney
Act concluded that the question always comes back to the question of
what the Congress intended--and they acknowledged that they did not
know the answer to this important question.  Since the McKinney Act
amendment was passed in 1992, a number of state agencies have
refunded bonds associated with Section 8 contracts issued between
1979 and 1984 and have not shared the savings with the government. 
While we agree with HUD that the savings at issue will generally be
smaller than the savings for the financing adjustment factor
contracts, which supported mortgage rates as high as 12 and 13
percent, we believe that HUD may have missed opportunities to provide
additional bond refunding savings to the government.  At the state
agencies we reviewed, mortgage rates supporting the refunded bonds
that may be subject to the McKinney Act varied, generally ranging
between 7.25 percent and 10 percent.  Furthermore, the bond refunding
savings for contracts other than financing adjustment factor
contracts that one state agency shares with the federal government
total about $14 million. 

      EXTENT TO WHICH STATES SHARE
      BOND REFUNDING SAVINGS
      VARIES
-------------------------------------------------------- Chapter 3:1.3

During the early to mid-1990s, state agencies were refunding the
bonds associated with Section 8 contracts to lower rates.  For
example, almost all of the bonds associated with contracts for
Section 8 project-based assistance at the five state agencies we
visited were refunded during this time.  We found that the five
agencies shared savings differently.  Specifically, the Illinois,
Maryland, and Massachusetts agencies generally share savings only for
financing adjustment factor contracts, while the Minnesota and New
Hampshire agencies share savings for other contracts as well. 

While none of the five state agencies shared savings for all
contracts entered into between 1979 and 1984ï¿½the time period
addressed in the October 1992 McKinney Act amendmentï¿½some agencies
shared savings only for financing adjustment factor contracts while
others shared savings for these and other contracts entered into
between 1979 and 1984.\2 For example, the Massachusetts agency shares
savings for its 35 financing adjustment factor contracts but does not
share savings for the other 70 contracts entered into between 1979
and 1984.  In contrast, the New Hampshire agency shares savings for
38 of the 48 contracts associated with refunded bonds that it entered
into between 1979 and 1984.  The 38 contracts include both financing
adjustment factor and other contracts. 

According to a HUD official, the agencies that shared savings for
both financing adjustment factor and other contracts entered into
between 1979 and 1984 may have believed they were required to share
savings under the McKinney Act.  We found, for example, that the
Minnesota agency determined that it was required to share savings
both for financing adjustment factor contracts and for contracts that
received special financing but did not explicitly require savings to
be returned to the government. 

However, other states concluded, independently of HUD, that the
McKinney Act generally was applicable only to financing adjustment
factor contracts.  For example, an official of the Maryland agency
said that, according to the agency's bond counsel, except for
financing adjustment factor contracts, the shared savings
requirements of the McKinney Act are applicable only when the Section
8 property mortgage is refinanced after a bond has been
refunded--that is, when the contract reduction method (see ch.  1) is
used.  A legal opinion from Maryland's bond counsel, issued in 1996
in conjunction with a bond refunding, discussed in detail the reasons
for believing that only financing adjustment factor contracts are
subject to the shared savings requirements of the McKinney Act.  In
concluding, the bond counsel noted that HUD does not have any formal
policy requiring the states to share savings for contracts that do
not have the financing adjustment factor.  The bond counsel said,
however, that HUD might in the future develop a policy for such
refundings, which would require savings to be shared and could be
applied retroactively.  While maintaining that the state agency would
likely prevail in a legal case on the issue, the bond counsel
acknowledged that a court could uphold such a claim by HUD. 

--------------------
\2 As discussed previously, HUD authorized higher interest rates and
Section 8 subsidies, but only the financing adjustment factor
contracts require agencies to refund the related bonds when interest
rates decline. 

   HUD'S GUIDANCE HAS NOT
   ELIMINATED EXCESS SECTION 8
   RENT INCREASES
---------------------------------------------------------- Chapter 3:2

HUD's general method for calculating automatic rent increases for
Section 8 properties does not consider the savings from bond
refundings.  To avoid increased costs to the government from
excessive rent increases to owners when bonds have been refunded, in
1997, HUD issued a notice requiring adjustments to the calculation of
rent increases for financing adjustment factor contracts.  However,
we found that the notice has several serious shortcomings, and
agencies' compliance with and understanding of this notice varies. 
As a result, the Section 8 program is incurring excess costs that it
could have avoided. 

      HUD'S RENT INCREASE NOTICE
      IS FLAWED
-------------------------------------------------------- Chapter 3:2.1

Issued on August 1, 1997, HUD Notice H 97-49 was intended to prevent
excessive rent increases following refundings of bonds associated
with financing adjustment factor contracts.\3 The notice establishes
a requirement for (1) calculating the contract rents that would have
resulted if the Section 8 contracts and mortgages had been reduced to
reflect the bond refunding savings and (2) applying the annual
Section 8 rent adjustment factor to this calculation.  However, the
notice does not explain the methodology for implementing the
procedure, nor does it provide an example of a calculation.  In
addition, the notice was issued for 1 yearï¿½until August 1998ï¿½and the
Department let it lapse without renewing it.  While HUD officials
told us that renewing the notice should not have been necessary
because the procedures apply to rent increases over the lives of the
Section 8 contracts, we do not believe the Department can hold its
field office staff and state and local agency staff accountable for
complying with the requirements of notices that are no longer in
effect.  Furthermore, we found that the Department's general guidance
on Section 8 rent increases, provided in notices 97-14 and 98-3,
contained no reference to the limit on rent increases contained in
notice H 97-49. 

The rent increase notice was issued 5 years after HUD's Inspector
General reported, in 1992, that Section 8 contracts were receiving
excessive rent increases after bonds were refunded and recommended
that the refunding savings be backed out before the rent increases
were calculated.  In 1993, the Department disagreed with the
Inspector General's recommendation, citing a HUD legal opinion
stating that HUD did not have the authority to limit Section 8
increases.  After the Inspector General reopened this recommendation
in 1997, the Department obtained a second legal opinion, which found
that HUD could limit the rent increases for financing adjustment
factor contracts. 

--------------------
\3 HUD Notice H 97-49 is titled Backing-Out Trustee Sweep Savings
Before Calculating AAFs for Projects Which Originally Received a FAF
and Whose Bonds Were Refunded. 

      AGENCIES' COMPLIANCE WITH
      HUD'S RENT INCREASE LIMIT
      VARIES
-------------------------------------------------------- Chapter 3:2.2

Since HUD issued its notice, two of the five state agencies we
visited have increased rents for financing adjustment factor Section
8 contracts.  We found that the Minnesota Housing Finance Agency
complied with the notice for contracts subject to the limitation
while the Massachusetts Housing Finance Agency did not. 

A Massachusetts Housing Finance Agency official told us in September
1998 that the agency did not adjust the rent increases because it had
not received the implementing information from HUD that the notice
said would be provided.  A HUD assessment of compliance with the rent
increase notice, initiated in Feburary 1999, also found that the
Massachusetts Housing Finance Agency was not adjusting the rent
increases for financing adjustment factor contracts.  The
Massachusetts agency responded to HUD that it had not complied with
the notice because (1) it was not included in the Department's rent
adjustment procedure notices, (2) it had expired in August 1998, and
(3) HUD had never provided the information required to implement the
notice.  An internal HUD review has recently identified other
agencies that did not comply with the notice when they processed rent
increases subject to the notice. 

Currently, the number of rent increases being processed for financing
adjustment factor contracts may be small because of other limits on
overall rent increases mandated by the Congress.  However, many of
the state agency contracts will be in effect for the next 15 to 20
years, and the rent increase limits are therefore likely to become
more applicable in the future. 

   HUD HAS NOT ENFORCED ITS
   PROHIBITION OF DUAL FEES
---------------------------------------------------------- Chapter 3:3

Even though HUD prohibits the collection of dual fees, some state
agencies have been collecting them for at least 6 years with HUD's
knowledge.  That is, some agencies have been collecting an annual
contributions contract fee and an override fee, although HUD's
regulations require state agencies to choose only one of these fees
for administering Section 8 contracts on behalf of the Department. 
As a result, HUD has added, at a minimum, tens of millions of dollars
to the Section 8 program's costs.  The Department is attempting to
resolve the issue for some of the agencies that have been collecting
dual fees for administering Section 8 contracts associated with
shared savings (refunding) agreements approved by HUD because of fee
information provided to the Department in this process.  However, we
found that HUD has not identified all state agencies that receive
dual fees and has not taken any actions when dual fees are being
collected for contracts that are not subject to shared savings
agreements with HUD.  Finally, HUD does not appear to have considered
options for recovering the override fee after it is no longer needed
to secure the bonds. 

      HUD'S REGULATION ALLOWS ONE
      OF TWO TYPES OF FEES, AND
      AGENCIES' CHARTERS LIMIT THE
      FEES' USE
-------------------------------------------------------- Chapter 3:3.1

State and local housing finance agencies that administer Section 8
contracts for HUD receive compensation for carrying out their
administrative responsibilities.  While all of the agencies are
eligible for an annual contributions contract fee, provided by HUD,
state agencies that finance Section 8 property mortgages with the
proceeds of state tax-exempt bond proceeds issued under the Internal
Revenue Service's (IRS) regulations may instead receive an override
fee (see ch.  1).  But whether an agency is eligible for one or both
fees, it is allowed to receive only one fee under a Section 8
regulation promulgated in 1980 (24 C.F.R.  883.606).\4

According to this regulation, a state housing finance agency that
chooses to collect an override fee cannot receive an annual
contributions contract fee.  HUD first prohibited dual fees in 1977
when it indicated that it would allow the state agencies to receive
an annual contributions contract fee only when the interest rates
that they were charging on their Section 8 property mortgages were
equal to the agencies' cost of borrowing.  HUD expected the state
agencies to take the annual contributions contract fee only in the
rare instances when they issued their bonds under HUD's regulations,
which do not permit an agency to charge more than its cost of
borrowing. 

According to the National Council of State Housing Agencies, the
charters and authorizing statutes of state agencies restrict their
spending to public purpose programs, including the costs of
administering those programs.  For example, the Minnesota agency said
that all income from any source is dedicated to providing affordable
housing and covering the operating costs necessary to fulfill its
mission.  Thus, the agency said, after covering its operating costs
and meeting its loan loss reserve requirements, it invests the
balance of its override fees in affordable housing activities. 
Similarly, the Maryland agency said that it uses its override fees to
pay the debt service on its bonds and the direct costs associated
with the bonds, such as the legal fees.  Each year, the agency
determines how much of the fees it can use for other purposes without
jeopardizing its bond rating.  The excess revenue may be used only
for affordable housing, community development, and budgeted
departmental operating costs. 

--------------------
\4 According to HUD officials, the regulation can be waived by the
Assistant Secretary for good cause and in writing because it is not a
statutory requirement. 

      HUD'S RESPONSE TO THE DUAL
      FEE ISSUE HAS BEEN SLOW AND
      INCONSISTENT
-------------------------------------------------------- Chapter 3:3.2

In 1992 and 1993, HUD's Inspector General reported that two state
agencies were receiving dual fees.  For example, the Inspector
General found that the Oregon state agency had received override fees
of $584,000 and annual contributions contract fees of $634,000 for
the same Section 8 contracts during the same period.\5 The Inspector
General recommended that HUD's Office of Housing (1) direct the field
offices to ensure that annual contributions contract fees are not
allowed when override fees are being collected and (2) determine the
amount of dual fees HUD had paid to agencies and require them to
return the overpayment to the Department.  In response to the
Inspector General's recommendations, HUD identified a number of state
agencies that were receiving dual fees and sent letters in August
1994 telling them to stop collecting dual fees unless they had
received a waiver of the regulation prohibiting dual fees.\6 The
Inspector General closed the recommendations on the basis of this
action. 

Six months later, however, in February 1995, HUD reversed its
position, stating that it would not enforce the dual fee prohibition
for Section 8 financing adjustment factor contracts associated with
bonds that had been refunded with the Department's approval.\7 This
letter said, however, that all bond-financing proposals submitted to
HUD in the future must comply with the dual fee prohibition. 
According to HUD officials, the Department changed its position after
some state agencies questioned the Department's legal authority to
enforce the dual fee prohibition.  The agencies pointed out that when
HUD negotiated shared savings agreements with them, it reviewed bond
refunding documents that showed the agencies would receive an
override fee.  The agencies maintain that the documents provided HUD
with the information needed to determine that the agencies were also
receiving annual contributions contract fees.  The agencies believe
that HUD therefore implicitly or explicitly approved the dual fees
for these contracts when it approved their refunding agreements. 

After finding that HUD had changed its position and was not going to
enforce the dual fee prohibition in existing cases, the Inspector
General reopened the recommendation.  Also, in October 1996, the
Inspector General requested that HUD not process any waivers before
the Inspector General completed a corrective action verification that
would be initiated the following month on the bond refunding audits. 

The debate between HUD and the Inspector General over the dual fee
issue has not been resolved.  While HUD has required the 17 dual fee
agencies it identified to request waivers of the dual fee regulation
and 14 agencies have done so, HUD has approved only one waiver
requestï¿½that of the Oregon Housing Finance Agency.  According to HUD
officials, the Department did not complete action on the remaining
waiver requests because in 1997 the Inspector General argued that the
dual fee agencies have created a valid debt to HUD and that under the
Debt Collection Act, HUD requires approval from the Department of
Justice to waive its regulation prohibiting dual fees.  In an October
1998 legal opinion, HUD's Office of the General Counsel did not
determine whether the payment of dual fees created a debt to HUD.  It
did agree that if a debt had been created, the Department would need
approval from the Department of Justice to waive the prohibition on
dual fees because of the requirements of the Debt Collection Act.\8

In November 1998, we met with the Deputy Assistant Secretary for
Housing and other officials to determine the status of HUD's actions. 
As a result of this meeting, the Deputy Assistant Secretary
established a team to resolve the issue; however, as of June 1999, it
had not been resolved, and HUD has yet to take action on the waiver
requests.  According to HUD, the program office (Housing) has asked
the Office of the General Counsel for advice on the options available
to the Department for resolving the issue. 

--------------------
\5 The dual fees cover 11 projects from 1984 through June 30, 1992. 
The annual override fee for most of this period was set at one-third
of the maximum allowable amount (0.50 percent) and was increased to
the maximum (1.50 percent) when the bonds were refunded in 1990.  The
Inspector General estimated that the higher override fee reduced the
bond savingsï¿½and thus increased the override fee--by $2.3 million. 

\6 We could not determine how many dual fee agencies HUD initially
identified and sent letters to.  However, a memorandum from the
National Council of State Housing Finance Agencies indicates that 14
agencies were to receive the letters.  Subsequently, in 1996, HUD
identified 17 agencies that it believed receives dual fees. 

\7 In 1996, HUD told agencies that the Department would have to
execute formal waivers of the dual fee prohibition and therefore
required them to request waivers of its regulation. 

\8 Under the Debt Collection Act of 1982, as amended, and
implementing regulations, the Department of Justice's concurrence is
required to terminate the collection of any claim exceeding $100,000. 

      HUD HAS NOT IDENTIFIED ALL
      CONTRACTS OR ALL AGENCIES
      WITH DUAL FEES
-------------------------------------------------------- Chapter 3:3.3

We found that HUD, in attempting to implement the Inspector General's
recommendation that it identify state agencies that receive dual
fees, has focused only on contracts that are subject to refunding
(shared savings) agreements with HUD.  However, we found that the
agencies also receive dual fees for Section 8 contracts that are not
subject to refunding agreements.  For example, four of the five state
agencies we visited received dual fees for 84 contracts with
refunding agreements and about 139 contracts with no refunding
agreements.  The bonds associated with most of these latter contracts
have been refunded, and the agencies receive override and
administrative fees.  According to HUD officials, if the Department
approves the agencies' requests for waivers of the dual fee
prohibition, the approvals would cover only the dual fees associated
with contracts that require the agencies to share savings with the
government under refunding agreements with HUD.  HUD would limit its
approval to these contracts because they alone provided the
Department with an opportunity to implicitly approve the dual fees. 

We also found that the Department had not identified all of the state
agencies that receive dual fees.  For example, although HUD had not
linked the Maryland agency with dual fees, we found that the agency
had received dual fees for 14 of its 35 Section 8 contracts.  We
confirmed that HUD had correctly identified the Massachusetts,
Minnesota, and New Hampshire agencies as dual fee agencies. 

In fiscal year 1998 alone, HUD paid these four state agencies over
$5.3 million in administrative fees, while the agencies were also
receiving override fees on the related bonds under the Section 8
program.  Because state agencies do not generally account for their
override fees on a project-by- project basis, we could not determine
the amounts of the related override fees associated with the Section
8 contracts for the four states.  However, we found that, in 1998,
one agency received override fees of close to $490,000 and
administrative fees of almost $423,000 for the same Section 8
contracts. 

Finally, an official at the Massachusetts agency---one of the dual
fees agencies identified by HUD that applied for a waiver in
1997ï¿½told us that his agency does not believe it is violating HUD's
dual fee prohibition.  The official stated that HUD's policy was
applicable only when a bond's original financing was in place.  The
agency maintains that since the bonds have been refunded, the
prohibition no longer applies.  Despite its position on the
legitimacy of receiving dual fees after bonds have been refunded, the
Massachusetts agency filed a waiver request with HUD in 1997 so that
it could continue to receive dual fees.  Moreover, the agency
received a letter from the Department in 1995 indicating that future
bond refunding proposals must comply with the dual fee prohibition.\9
Finally, in 1996 in the Code of Federal Regulations, HUD reiterated
its policy that the dual fee prohibition applies to bond refundings
(24 C.F.R.  811.110 (b)). 

--------------------
\9 The 1995 letter said that HUD would not require compliance with
the dual fee prohibition when HUD had approved refunding agreements. 
As discussed above, in 1996, HUD required the agencies to request
waivers of the regulation. 

      ONE APPROVED WAIVER AND
      PENDING WAIVER REQUESTS RELY
      IN PART ON PLEDGES OF FEES
      AS SECURITY FOR BONDS
-------------------------------------------------------- Chapter 3:3.4

In June 1997, the Assistant Secretary for Housing-Federal Housing
Commissioner approved the Oregon Housing and Community Service
Department's request for a waiver of the dual fee prohibition,
relying in large part on the Oregon agency's pledges and uses of fee
income.  According to the letter approving the waiver, the Oregon
agency has pledged both the override and the contract administration
fees as security for the bonds and this pledge is contained in
official bond documents.  The letter said that HUD did not wish to
upset the existing security arrangements relied on by bond rating
agencies and purchasers of the bonds.  The letter also recognized
that the Oregon agency had allocated the override fee to several
affordable housing programs.  The Assistant Secretary also stated
that in collecting duplicate fees, the Oregon agency relied on
decisions HUD made in processing and approving the bond refunding
transactions. 

Most of the 14 waiver requests submitted by state agencies also state
that the documents provided to HUD for shared savings agreements
indicated that the state agencies would continue to collect both the
override fees permitted by the federal tax laws and the Section 8
administrative fees.  Many of the waiver request letters, including
the Oregon letter, indicated that the override fee was needed to
support state housing programs but was also restricted because it was
pledged as security to the bondholders. 

From discussions with state agencies, we understand that an override
fee may be restricted for a period of time but may become available
to an agency while a bond is still outstanding.  Some states may use
their override fees for their housing programs or their expenses,
while others may keep the fees in their bond accounts, investing them
to earn investment income.  The fee amounts remaining in the bond
accounts accrue to the agencies when the bonds are paid off.  In
1997, at least one HUD official asked whether the dual fee issue
could be resolved by allowing the state agencies to keep the override
fees while required by bond pledges and then require the agencies to
return the fees to HUD.  The Department does not appear to have
pursued this suggestion for resolving the dual fee problem. 

According to a 1997 study by a bond rating agency, state agencies
have built up equity over the last two decades in bond accounts
associated with Section 8 contracts primarily by accumulating
override fees.\10 Thus, the financial outlook for state agencies'
Section 8 bonds was expected to remain very strong, in large part
because the agencies have sufficient equity available in the event of
significant increases in nonperforming loans.  For example, the study
indicated that agencies typically had equity in excess of 12 percent
of the bonds outstanding.  One agency's equity ($87 million) was
equal to 56 percent of its outstanding bonds ($154 million).  The
state agencies will generally continue to receive the override fees
annually throughout the terms of their Section 8 contracts, many of
which have another 15 to 20 years remaining. 

--------------------
\10 Bonds Secured by Section 8 Subsidies ï¿½ A Tale of Two Outlooks,
Moody's Investor Service, November 1997. 

   CONCLUSIONS
---------------------------------------------------------- Chapter 3:4

For more than 6 years, HUD has missed opportunities to protect the
federal government's interests and to reduce Section 8 rental
assistance costs: 

  -- Because HUD has not resolved the applicability of the October
     1992 amendment of the McKinney Act to state agencies, the
     Department may have missed opportunities for the government to
     obtain certain bond refunding savings.  If HUD determines that
     this provision is applicable to state agencies and that it can
     begin to require them to share the bond refunding savings, these
     savings will reduce the Section 8 costs that the federal
     government will bear for many years to come. 

  -- By taking so long to publish procedures for limiting Section 8
     rent increases after bonds have been refunded, HUD missed
     opportunities to lower the Section 8 program's long-term costs. 
     Moreover, by failing to ensure compliance with the procedures
     after publishing them in a notice, allowing the notice to expire
     after 1 year, and not providing promised guidance or examples of
     calculations for implementing the procedures, HUD missed further
     opportunities to control the program's costs. 

  -- Finally, by failing to enforce its regulation prohibiting dual
     fees, HUD has added tens of millions of dollars to the Section 8
     program's costs.  If allowed to continue, the fee payments will
     unnecessarily increase the program's costs for the next 15 to 20
     years.  In the meantime, a bond agency's data indicate that the
     state agencies are accruing significant amounts of equity from
     override fees that will be fully available to the agencies when
     the bonds are paid off, if not sooner.  In addition, HUD has not
     identified all of the contracts or all of the state agencies
     receiving dual fees.  Because it did not recognize that some
     agencies were receiving dual fees for contracts that are not
     subject to refunding agreements approved by HUD, the Department
     missed opportunities to enforce the dual fee regulation in these
     cases where, in the opinion of HUD officials, the agencies do
     not have a basis for requesting waivers.  Furthermore, Housing
     officials do not appear to have considered options other than
     waiving the dual fee prohibition.  One such option would allow
     the agencies to continue receiving both fees until the
     documented bond commitments have been satisfied and then
     requiring the agencies to return the fees to HUD.  In this way,
     the government could limit the impact of the excess fee costs on
     the Section 8 program. 

   RECOMMENDATIONS
---------------------------------------------------------- Chapter 3:5

We recommend that the Secretary of Housing and Urban Development
determine whether the state agencies are required, under the McKinney
Act, to share savings from refunding bonds associated with all
contracts entered into between 1979 and 1984.  If so, the Department
should revise its applicable rules and regulations to clarify the
requirements for sharing bond refunding savings with the federal
government.  For contracts associated with bonds that have already
been refunded, HUD should determine whether it can require the state
agencies to begin sharing the Section 8 savings they currently
retain. 

To ensure that state and local housing finance agencies comply with
HUD's guidance on deducting bond refunding savings before calculating
rent increases, we recommend that the Secretary require the
Department to provide the state and local agencies with the
appropriate methodology and examples of calculations and ensure that
the rule is kept current and integrated into the Department's
guidance on annual Section 8 rent increases. 

To eliminate excess costs from paying dual fees to state agencies
administering Section 8 rental assistance contracts, we recommend
that the Secretary require the Department to enforce its regulation
prohibiting dual fees unless there is a documented, sound, and
equitable basis for waiving the regulation.  To enforce the
regulation, the Department should identify all Section 8 contracts
for which state agencies receive both an administrative fee under the
Section 8 contract and an override fee, including those contracts
that are not subject to a refunding agreement with HUD. 

   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 3:6

In commenting on the draft report's discussion and recommendation
concerning the applicability of the McKinney Act's shared savings
provision (section 1012, as amended in October 1992) to state
agencies, HUD says that the draft report and recommendation direct
the Department to take action where its legal authority is unclear. 
In fact, the draft report recommended that HUD clarify its legal
authority (determine whether the state agencies are required, under
the McKinney Act, to share savings from refunding certain bonds with
the government) and revise its rules and regulations accordingly. 
However, we recognize that a statement in the executive summary may
have caused some confusion, and we have revised it for clarity and
greater consistency with our discussion in the body of the report. 
Specifically, the draft executive summary said that HUD has not
issued guidance to the state agencies directing them to comply with
the October 1992 amendment to section 1012 of the McKinney Act, which
requires the agencies to share certain bond refunding savings with
the government.  As is consistent with the discussion in this
chapter, the final report states that HUD has not issued guidance to
the state agencies on how to comply with the October 1992 amendment
to section 1012 of the McKinney Act, which provides for the agencies
to share certain bond refunding savings with the federal government. 

Although HUD has not resolved the applicability of the McKinney Act
amendment to state agencies, HUD's previous actions indicated that
the Department believed it could require state agencies to share
these savings.  For example, the comments do not recognize that (1)
the Department's Office of the General Counsel concluded that HUD
could require state agencies to comply with the shared savings
provision and (2) HUD attempted in 1996 to require the agencies, by
regulation, to share these savings.  Today, more than 6 years after
the shared savings provision was amended, HUD has not clarified the
provision's applicability to the state agencies.  In the absence of a
clear interpretation, some state agencies have retained the bond
refunding savings, while others have shared the savings with the
government.  Thus, HUD's inaction may have deprived the federal
government of opportunities to share savings and reduce its Section 8
costs. 

In further commenting on this recommendation, HUD says that we
recommend that the Department retroactively recover the portion of
the bond refunding savings that some state agencies have retained in
their entirety.  In fact, our recommendation is prospective, stating
that if HUD finds the agencies are required to share their savings,
then it should determine whether it can require them to begin sharing
the savings they currently retain. 

In a final comment on this recommendation, HUD questions whether data
exist to support our statement that, without clarification of the
McKinney Act's shared savings provision, state agencies may retain
ï¿½tens of millionsï¿½ of dollars that they could be legally required to
share with the government.  This figure, cited to convey the
magnitude of the savings at issue, was based primarily on information
we had obtained on Section 8 bond refundings by state agencies with
and without shared savings agreements.  It also considered the bond
refundings savings, amounting to millions of dollars annually, that
two state agencies in our review share with the government for
contracts other than financing adjustment factor contracts. 
Nonetheless, we agree with HUD that the savings for contracts other
than financing adjustment factor contracts will generally be smaller
than the savings for financing adjustment factor contracts, which
supported mortgage rates as high as 12 and 13 percent.  For example,
as indicated in the final report, the mortgage rates for contracts
other than financing adjustment factor contracts at the state
agencies we reviewed generally range between 7.25 percent and 10
percent.  We revised the final report to state that, in our view, HUD
may have missed opportunities to provide additional bond refunding
savings to the government to offset the related Section 8 costs, and
we deleted the estimate of potential savings. 

In response to our recommendation that HUD ensure the currency of its
rule on deducting bond refunding savings before calculating rent
increases and integrate this rule into its guidance on annual Section
8 rent increases, HUD agreed to reissue its expired notice and to
evaluate the need for cross-referencing other Section 8 rent
directives.  HUD also agreed to ensure compliance with its
rent-adjustment rule by providing state and local agencies and HUD
field staff with clarifying instructions for compliance with the
notice. 

In commenting on our recommendation that the Department enforce its
regulation prohibiting dual fees unless there is a documented, sound,
and equitable basis for waiving the regulation, HUD said that if the
refunding agreements it approved had been accompanied by executed
waivers of the dual fee prohibition, no dispute about dual fees would
have arisen.  We agree with HUD that waivers, if executed in
compliance with the Department's waiver rules and applicable federal
laws, would have avoided the issue.  However, waivers have not been
executed, and the issue remains unresolved. 

HUD also suggests that we consider the dual fee as an incentive that
encouraged state agencies to refund the bonds associated with their
Section 8 contracts and, hence, provide bond refunding savings to the
government.  According to HUD's estimate, these savings will exceed
$150 million over the lives of the contracts.  Whether these savings
will offset the costs of the dual fees is unknown because neither the
number of state agencies that take dual fees nor the number of
contracts associated with dual fees is known.  Nevertheless, we
estimated in our draft report that dual fees add tens of millions of
dollars to the Section 8 program's costs.  We based this estimate on
the fees paid in 1998 to four state agencies for administering
contracts that are also associated with override fees.  As we
indicated in the draft report, these fees totaled $5.3 million for 1
year.  Most state agencies' contracts will be in effect for another
10 to 20 years, and administrative fees, which are tied to unit
costs, are likely to increase.  Thus, the dual fees for these four
agencies alone will add tens of millions of dollars to the Section 8
program's costs.  Furthermore, the draft report indicates that 14
state agencies have requested waivers of the dual fee prohibition and
shows that HUD has not identified all state agencies that receive
dual fees.  Thus, we believe our estimate of tens of millions of
dollars in dual fee costs is conservative.  We revised the final
report to say that the dual fees have added, at a minimum, tens of
millions of dollars to the Section 8 program's costs. 

Finally, HUD does not indicate in its comments how it will identify
the state agencies that receive dual fees for Section 8 contracts
that are not subject to refunding agreements approved by HUD.  As
discussed in the report, HUD officials said that, for these
contracts, they had no basis for approving waivers of the dual fee
regulation.  We urge the Department to respond expeditiously so that
these fees do not continue to burden the Section 8 program. 

In commenting on the draft report, the National Council of State
Housing Agencies disagreed withï¿½but misstatedï¿½our conclusion and
recommendation about the McKinney Act's requirements.  According to
the Council, we wrongly concluded that HUD has the authority to
require state agencies to share certain bond refunding savings with
the federal government.  In fact, the draft report recommended that
the Secretary determine whether the state agencies are required,
under the McKinney Act, to share certain bond refunding savings. 
This recommendation is based on our conclusion that HUD has not
provided the agencies with a clear interpretation of the McKinney
Act's shared savings provision (section 1012, as amended in October
1992), and, as a result, some state agencies are sharing certain bond
refunding savings with the government while others are not.  The
Council also disagreed with the report for not recognizing that HUD
effectively waived its prohibition of dual fees by approving
refunding transactions under which state agencies receive both fees. 
As the draft report clearly states, this is the position of the state
agencies.  However, HUD is required to issue formal waivers when it
does not enforce a regulation such as the dual fee prohibition. 

HUD'S INFORMATION ON THE PHYSICAL
AND FINANCIAL CONDITION OF THE
UNINSURED PORTFOLIO IS LIMITED,
BUT 10 STATE AGENCIES REPORTED FEW
PROBLEMS IN THEIR PORTFOLIOS
============================================================ Chapter 4

To monitor the physical and financial condition of properties in the
uninsured portfolio, HUD requires annual physical inspections and
generally requires annual audited financial statements.  The 10 state
agencies we surveyed reported having monitoring policies and
procedures in place that conformed to HUD's guidance.  As of December
1998, HUD had limited information on the physical and financial
condition of the uninsured portfolio.  According to the Department's
central database, the majority of the properties were in satisfactory
or superior physical condition; however, these ratings were not based
on objective criteria, and their reliability was therefore unknown. 
The 10 state agencies reported that nearly all of the properties in
their uninsured Section 8 portfolios were in satisfactory or better
physical condition, yet their ratings, like HUD's, were subjective. 
Also as of December 1998, HUD's database contained no information on
the financial condition of properties in the uninsured portfolio. 
According to the 10 state agencies, a very small fraction of their
properties (under 4 percent) warranted special monitoring because of
financial or other problems.  In mid-1998, HUD began to establish
centralized procedures, including objective rating criteria, to
improve the monitoring of multifamily properties in its uninsured and
other portfolios. 

   HUD AND STATE AGENCIES MONITOR
   CONDITIONS TO ENSURE COMPLIANCE
   WITH FEDERAL REQUIREMENTS
---------------------------------------------------------- Chapter 4:1

HUD field offices currently administer about 80 percent of the
uninsured Section 8 housing assistance contracts, while state or
local housing finance agencies generally administer the balance. 
Under their contracts with HUD, the state and local agencies are
responsible for monitoring the physical and financial condition of
the Section 8 properties in their portfolios, as well as for
evaluating other aspects of the properties' management, such as their
leasing and occupancy procedures.  The purpose of this monitoring is
to ensure that property owners and managers comply with the terms of
their Section 8 subsidy contracts, regulatory agreements, and other
administrative requirements. 

HUD requires annual physical inspections of properties in the
uninsured portfolio.  According to HUD's handbooks, HUD field staff,
HUD contractors, or state or local contract administrators should
visit Section 8 properties at least annually to perform physical
inspections and to determine whether property owners are providing
decent, safe, and sanitary housing.  The physical inspections are to
include examinations of buildings, grounds, and mechanical systems. 
The 10 state agencies we surveyed reported having policies and
procedures in place for performing the minimum number of physical
inspections that HUD requires.  In addition, most of the agencies
reported using HUD's inspection forms to perform these inspections. 
The agencies also said they usually increase the frequency of their
inspections when they identify unsatisfactory conditions. 

HUD requires most property owners to submit annual audited financial
statements to HUD or its contract administrators, who are then
required to perform oversight reviews.\1 Besides reviewing the
required financial audits, state agencies reported using other
procedures to monitor their properties, such as examining operating
reports, approving budgets and rent increases, processing requests
for using cash reserves, analyzing monthly financial reports, and
making on-site management assessments.  In addition, each agency
identified methods it used to strengthen its monitoring of properties
with operating difficulties, such as financial instability, physical
and maintenance problems, and management or ownership concerns. 

--------------------
\1 According to HUD officials, property owners who executed Section 8
contracts with HUD before 1980 may not be required to submit audited
financial statements to HUD or its contract administrators. 

   INFORMATION ON THE PHYSICAL AND
   FINANCIAL CONDITION OF
   UNINSURED SECTION 8 PROPERTIES
   WAS LIMITED
---------------------------------------------------------- Chapter 4:2

Information on the physical and financial condition of uninsured
Section 8 properties was limited.  We obtained information on the
uninsured Section 8 portfolio as a whole from HUD's central database. 
However, the information in this database on the properties' physical
condition was subject to several limitations:  Inspection results
were reported for only 63 percent of the properties, 42 percent of
the inspections were at least 3 years old, and the ratings assigned
by inspectors were not based on objective criteria.  The database
contained no information on the results of the annual financial
audits that HUD requires.  While the information on the physical
condition of properties in the 10 state agencies' uninsured Section 8
portfolios was more recentï¿½based primarily on inspections performed
in 1998--the ratings were also subjective because they were not based
on objective criteria.  Five of the state agencies rated the
financial condition of their properties, and seven agencies rated the
overall condition of their properties, taking into account their
management as well as their physical and financial condition.  The
information provided by the state agencies is presented in more
detail in appendix IV. 

The 10 state agencies we surveyed were responsible for monitoring
1,167 of the 2,278 properties that, as of December 1998, were in the
uninsured Section 8 portfolios administered by 30 state agencies
nationwide.\2 About 43 percent of the 1,167 properties were designed
specifically for the elderly (41 percent) or the disabled (2
percent), and 57 percent of the properties were designed for families
or families and the elderly. 

--------------------
\2 The state agencies may also administer other Section 8
project-based contracts associated with properties that were not
included in the scope of this review, such as FHA-insured properties
with project-based assistance or moderate rehabilitation properties. 

      MOST PROPERTIES WERE
      REPORTED IN GOOD PHYSICAL
      CONDITION, BUT THE RATINGS
      ASSIGNED DURING INSPECTIONS
      WERE OF LIMITED USE
-------------------------------------------------------- Chapter 4:2.1

HUD's information on the physical condition of properties in the
uninsured portfolio was limited.  As of December 1998, HUD's central
database contained information on the results of inspections
performed at 7,867 properties, or 63 percent of 12,488 properties in
the uninsured portfolio.  The proportion of inspections reported for
the uninsured programs ranged from 29 percent for the
elderly/disabled capital advance program to 89 percent for the loan
management set aside program.  However, for 42 percent of the
uninsured properties, inspections had not been performed within the
past 3 years.  According to the ratings assigned during these
inspections, 92 percent of the 7,867 properties were in
ï¿½satisfactoryï¿½ or ï¿½superiorï¿½ physical condition, while 6 percent were
in ï¿½below averageï¿½ or ï¿½unsatisfactoryï¿½ condition.  However, these
data have limitations because, until very recently, HUD lacked
uniform standards for reporting the physical condition of its
multifamily portfolio.  Without such standards, property ratings were
assigned subjectively, making comparisons among properties, field
locations, and programs difficult and questionable.  HUD recognizes
that without uniform and objective physical inspection standards, the
data are of limited use in overseeing the portfolio.  As discussed
later this chapter, the Department has developed new property
inspection standards designed to address the shortcomings of the
prior inspection standards. 

The 10 state agencies we surveyed reported that 95 percent of the
properties in their portfolios were in satisfactory or superior
physical condition.  (See fig.  4.1).  These properties represent 97
percent of the apartment units in their portfolios.  While all of the
agencies used HUD's rating scale of superior, satisfactory, below
average, or unsatisfactory to classify the physical condition of
their properties, they used their own criteria and judgment to assign
the ratings.  As a result, the ratings are difficult to compare and
their reliability is unknown.  (See app.  IV, table IV.2, for a
state-by-state summary of the properties and units in each rating
category). 

   Figure 4.1:  Physical Condition
   of Properties, as Reported by
   10 State Agencies

   (See figure in printed
   edition.)

Note:  Information is based primarily on the results of the most
recent (1998) physical inspections. 

Source:  GAO's analysis of information provided by the 10 state
agencies. 

      HUD'S DATABASE HAS NO
      INFORMATION ON THE RESULTS
      OF FINANCIAL AUDITS, BUT
      STATE AGENCIES REPORTED FEW
      FINANCIAL PROBLEMS
-------------------------------------------------------- Chapter 4:2.2

Even though HUD requires property owners to submit annual audited
financial statements, HUD's central database contained no information
on the financial condition of the uninsured portfolio.  As of
December 1998, the database included dates relevant to the audited
financial statements, such as when the audits were performed or when
HUD received the statements.  But the database did not contain the
results of the audits.  As a result, no conclusions can be drawn
about the financial status of the uninsured portfolio. 

HUD does not require the state agencies to assign a rating to the
financial condition of their properties.  Nevertheless, 5 of the 10
agencies we surveyedï¿½California, Michigan, New Hampshire, Tennessee,
and Wisconsinï¿½had rated the financial condition of their properties. 
As shown in figure 4.2, these five states reported that about 96
percent of the 474 properties in their portfolios were in
satisfactory, good, excellent, or superior financial condition while
about 4 percent were in below average, poor, or unsatisfactory
condition.  (See app.  IV, table IV.3 for a state-by-state summary of
financial condition ratings.) Some of the underlying factors the
agencies considered in evaluating a property's financial stability
included cash reserve levels, cash flow, debt service coverage,
operating expenses, and liquidity. 

   Figure 4.2:  Financial
   Condition of Properties, as
   Rated by Five State Agencies

   (See figure in printed
   edition.)

Source:  GAO analysis of information provided by the California,
Michigan, New Hampshire, Tennessee, and Wisconsin housing finance
agencies. 

      ACCORDING TO STATE AGENCIES,
      FEW PROPERTIES HAVE OVERALL
      OPERATING PROBLEMS OR
      REQUIRE SPECIAL MONITORING
-------------------------------------------------------- Chapter 4:2.3

Though not required by HUD, ratings of the overall condition of their
properties were provided by 7 of the 10 state agencies we surveyed. 
To determine the overall condition of their properties, the state
agencies generally considered the properties' management and physical
and financial condition.\3

As shown in figure 4.3, the seven agencies reported that 95 percent
of the 666 properties in their portfolios were in satisfactory or
better overall condition and that only 5 percent were in below
average or unsatisfactory condition.  According to the National
Council of State Housing Agencies, the healthy overall physical and
financial condition of the properties in the state agency portfolio
is largely attributable to the successful asset management and
oversight practices employed by the state agencies.  (See app IV,
table IV.4, for a state-by-state summary of overall ratings). 

   Figure 4.3:  Overall Condition
   of Properties, as Rated by
   Seven State Agencies

   (See figure in printed
   edition.)

Source:  GAO's analysis of the overall condition ratings provided by
the California, Illinois, Maryland, Massachusetts, Michigan, New
Hampshire, and Tennessee housing finance agencies. 

In addition, the 10 state agencies reported that only 45--or fewer
than 4 percentï¿½of the 1,167 properties in their portfolios warranted
special or extra monitoring attention.  The reasons cited for
assigning properties to the special monitoring category included
financial problems, deferred maintenance, occupancy issues, and/or
concerns over management or ownership.  Special monitoring usually
involved increased oversight, such as additional site visits and
evaluations; additional financial or operational reporting
requirements; or specific workout agreements.  (See app.  IV, table
IV.5, for data on each state's special monitoring.)

--------------------
\3 The Minnesota and Oregon agencies did not provide either financial
or overall ratings for the 350 properties in their portfolios because
their monitoring systems did not collect information in ways that
would allow them to do so. 

   HUD IS TAKING ACTION TO OBTAIN
   MORE RELIABLE AND COMPLETE DATA
   ON ITS PORTFOLIOS
---------------------------------------------------------- Chapter 4:3

HUD is taking steps to improve the reliability and availability of
its data on the multifamily properties in its various portfolios,
including both the uninsured and insured Section 8 portfolios.  In
September 1998, HUD formed the Real Estate Assessment Center in
Washington, D.C., to improve its processes for monitoring the
physical and financial condition of its properties.  The Center has
central responsibility for accumulating and analyzing the results of
(1) the physical inspections of HUD's properties nationwide and (2)
the annual audited financial statements for HUD's entire portfolio. 
Multifamily asset managers will then use the results of these
analyses to develop oversight strategies for individual properties. 

In the past, HUD lacked an objective system for classifying the
results of the physical inspections it requires for its multifamily
uninsured and insured properties.  The Real Estate Assessment Center
developed such a system.  The system uses a point scale to translate
the results of an on-site evaluation into an overall, quantitative
rating for each property.  HUD has acquired and trained contractors
to perform physical inspections using the Center's new inspection
protocol.  By the end of 1999, the Department expects to have
baseline data on over 30,000 multifamily properties, including
uninsured and insured Section 8 properties and insured properties
that do not receive Section 8 assistance. 

At the end of April 1999, the contractors had completed physical
inspections at 4,513 properties under various multifamily housing
programs.  Of 2,401 uninsured properties that were inspected, 39
percent were in excellent condition, 46 percent were in good
condition, 13 percent were in fair condition, and 2 percent were in
poor condition.\4 (See app.  V, table V.1 for additional information
on the results of the inspections.)

The Real Estate Assessment Center has also developed a central system
for analyzing financial information from the annual audited financial
statements that the owners of multifamily properties are required to
submit under the terms of their Section 8 housing assistance payment
contracts.  The Center will compare this financial information with
specific performance ratios, compliance indicators, and other
standards and benchmarks for individual portfolios and for the
industry as a whole.  HUD expects these comparisons to provide
information that its field offices will find useful in their
monitoring and enforcement efforts.  In general, the owners of
properties in the rural housing and state agency programs are not
required to report to the Center because they do not submit their
annual audited financial statements to HUD.  The Center required all
other owners to submit their 1998 audited financial statement
information by June 30, 1999. 

--------------------
\4 About 61 percent the 2,401 inspections were performed at
properties in the elderly/disabled loan portfolio, while the
remainder were performed at other uninsured properties.  The Real
Estate Assessment Center was unable to provide a breakdown showing
how many of the remaining properties were in the state agency
portfolio of 2,278 Section 8 properties. 

   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 4:4

According to the National Council of State Housing Agencies, our
draft report did not adequately stress the connection between the
superior physical and financial condition of uninsured properties in
the state agency program and the state agencies' management and
oversight.  However, the scope of our review was not sufficient to
link the condition of the properties to the state agencies'
management and oversight.  Specifically, to satisfy the mandate, we
agreed to determine how the state agencies oversee the Section 8
properties in their portfolios and to provide information on the
physical and financial condition of the properties in 10 state
agencies' portfolios.  These properties account for about 50 percent
of the properties in the state agency program.  However, we included
in the final report the Council's opinion that the physical and
financial health of the properties in the state agency portfolio is
largely attributable to the successful asset management and oversight
practices employed by the state agencies. 

INFORMATION ON THE SIZE OF THE
UNINSURED SECTION 8 PORTFOLIO
=========================================================== Appendix I

This table provides additional information on the number of
properties, contracts, and assisted units in the uninsured portfolio,
broken out by the eight programs making up this portfolio. 

                Table I.1 Number of Uninsured Section 8
                  Properties, Contracts, and Units, by
                      Program, as of December 1998

                                                                 Total
                                                              assisted
Program                         Properties     Contracts         units
----------------------------  ------------  ------------  ------------
Elderly/disabled loan                4,446         4,451       209,178
State agency                         2,278          2280       172,794
New construction/                    1,678         1,708        91,435
 substantial rehabilitation
Rural housing                        1,518         1,525        44,788
Elderly/disabled capital             1,510         1,510        38,449
 advance
Loan management set-aside              513           678        32,674
Multifamily property                   521           530        39,829
 disposition
Housing preservation                    24            26         3,069
======================================================================
Total                               12,488        12,708       632,216
----------------------------------------------------------------------
Source:  HUD's Real Estate Management System (REMS) and Section 8
expiring contracts databases, as of Dec.  1998. 

INFORMATION ON THE HUD DATABASES
USED TO ANALYZE THE SECTION 8
UNINSURED PORTFOLIO
========================================================== Appendix II

This appendix provides more detailed information on the Department of
Housing and Urban Development's (HUD) databases we used to determine
what information HUD has on the Section 8 assistance provided to, and
the physical and financial condition of, the Section 8 properties
that are not insured by HUD's Federal Housing Administration (FHA). 

To develop information on the Section 8 assistance provided to
uninsured properties, including data on expenditures, rents, and
expiring contracts, we obtained three database files from HUDï¿½two
from the Office of Housing and one from the Office of Chief Financial
Officer.  The first file from Housing, a database on Section 8
contracts expiring through the year 2038, included all insured and
uninsured contracts receiving Section 8 project-based assistance as
of December 1998.  The second file, a data extract from HUD's Real
Estate Management System (REMS)ï¿½the Department's official source of
information on multifamily housingï¿½contained information on all
active uninsured properties as of December 24, 1998, including
information on the various programs that make up HUD's uninsured
Section 8 portfolio.  The file from the Office of Chief Financial
Officer contained fiscal year 1998 expenditure and receipt data for
all Section 8 contracts from the Department's Program Accounting
System (PAS).  These three database files provided us with the
information found in chapters 1 and 2. 

To determine the expenditures and per-unit costs associated with
uninsured Section 8 contracts, we used PAS to calculate the net
expenditures (expenditures minus receipts) during fiscal year 1998
for each of the eight primary programs in the portfolio.  We then
calculated the average per-unit subsidy cost for each program.  After
comparing the results of the two sets of calculations to identify any
significant variances, we met with HUD officials to discuss the
reasons for, and the factors potentially contributing to, those we
found. 

We used HUD's database of expiring Section 8 contracts and the data
extract from REMS to provide information on the Section 8 contract
and fair market rents (FMR) for properties in the uninsured
portfolio.  To do this, we matched the universe of 12,708 uninsured
contracts from REMS with corresponding information on the contract
and fair market rents from the expiring Section 8 contracts database. 
For each of the eight programs, we then determined the percentage of
units whose contract rents exceeded the fair market rents for their
area.  We also used the expiring contracts database to identify the
expiration dates associated with project-based contracts in the
uninsured portfolio.  We then analyzed each contract's expiration
date to identify variances among the eight programs in the uninsured
portfolio. 

To ascertain the physical and financial condition of the uninsured
properties, we matched the universe of uninsured properties in the
data extract from REMS with source tables in REMS that contained
physical and financial information.  We included in our analysis only
the results of the most recent physical and financial reviews,
excluding the results of older reviews included in the tables for
some properties.  We then aged these results for each of the programs
in the portfolio.  Although REMS contained rating data from the
physical inspections of many of the uninsured properties in the
portfolio, it did not contain corresponding information from the
financial reviews. 

To ensure the reliability of the information in this report, we met
with HUD officials to discuss various data issues and reviewed
selected data fields in each of the three files for completeness,
accuracy, and relevance.  For example, we found that some recently
renewed contracts had not yet been included in the universe of active
uninsured contracts.  We discussed this issue with Housing officials,
who developed a methodology to ensure that these contracts were
included in the active uninsured portfolio.  We also monitored
efforts by the Multifamily Data Quality Task Force to identify,
measure, and fully verify ï¿½criticalï¿½ Section 8 data elements in
REMSï¿½some of which are used in this report.  Such elements include
data on financing, units, and contract rents in the uninsured
portfolio.  To check the expenditure and per-unit cost information
for accuracy and completeness, we matched the universe of active
uninsured Section 8 contracts with the PAS expenditure data and
tested the key data elements to identify data that were missing or
inaccurate.  In 1997, HUD reported that PAS was in compliance with
the Federal Manager's Financial Integrity Act (FIA).  In addition,
HUD's Office of Inspector General examined funding and expenditure
data as part of its financial audit for fiscal years 1996-97 and did
not identify data errors associated with Section 8 contract
expenditures that were material to HUD's financial statements.  We
also reviewed the Department's methodology for developing the ratios
of contract rents to fair market rents.  On the basis of these
reviews, we determined that the data were sufficiently reliable to
use in this study.  We made some minor adjustmentsï¿½affecting 75
contractsï¿½to the universe of active uninsured Section 8 contracts. 
We discussed these changes with HUD officials, who agreed that our
adjustments were reasonable. 

INFORMATION ON SECTION 8 CONTRACT
RENTS AND EXPIRATIONS
========================================================= Appendix III

This appendix provides more detailed information on (1) Section 8
contract rents in relation to HUD's fair market rents and (2) Section
8 contract expirations for the uninsured portfolio. 

Table III.1 relates to table 2.2 in chapter 2, which identifies, for
each of the programs in HUD's uninsured Section 8 portfolio, the
percentage of assisted Section 8 units whose Section 8 rents exceed
HUD's fair market rents.  This table breaks out the overall
percentage into four other categories, which show the extent to which
the rents exceed the fair market rents. 

                                       Table III.1
                         
                            Percentage of Assisted Units Whose
                          Section 8 Contract Rents Exceed HUD's
                         Fair Market Rents for Programs in HUD's
                           Uninsured Section 8 Portfolio, as of
                                      December 1998

                                         Percentage of assisted units
                                   ----------------------------------------
                       With rents    With rents    With rents
                      between 101   between 121   between 141    With rents    Total with
                          and 120       and 140       and 160      over 160         rents
                       percent of    percent of    percent of    percent of     exceeding
                      fair market   fair market   fair market   fair market   fair market
Program                     rents         rents         rents         rents         rents
------------------  -------------  ------------  ------------  ------------  ------------
State agency                   17            24            23            28            91
Elderly/disabled               19            23            20            25            88
 loan
Rural housing                  20            31            23            12            86
New construction/              29            22            14             8            72
 substantial
 rehabilitation
Preservation                   27            12             4             0            42
Loan management                19             3             0             1            24
 set-aside
Property                       17             3             1             0            21
 disposition
Elderly/disabled                4             1             0             0             5
 capital advance
-----------------------------------------------------------------------------------------
Note:  These data cover 12,288 of the 12,708 contracts in HUD's
uninsured portfolio for which data on rents and HUD's fair market
rents were available. 

Source:  HUD's Section 8 Expiring Contracts Database, Dec.  1998. 

Table III.2 relates to figure 2.3 in chapter 2, which shows when
uninsured Section 8 contracts will expire.  This table breaks out the
expiration dates for contracts in each of the programs that make up
HUD's uninsured Section 8 portfolio. 

                  Table III.2 Percentage of Contracts
                 Expiring Within Indicated Fiscal Year
                Ranges, by Program, as of December 1998

Program                  1999-2004   2005-2009   2010-2014   2015-2021
----------------------  ----------  ----------  ----------  ----------
Loan management set-            99           0           0           0
 aside
Rural housing                   97           2           0           2
Preservation                    92           0           0           0
Property disposition            58          29          12           0
New construction/               48          16          25          10
 substantial
 rehabilitation
Elderly/disabled loan           41          33          25           1
Elderly/disabled                15           4          16          62
 capital advance
State agency                     6           8          41          42
----------------------------------------------------------------------
Source:  GAO's analysis of HUD's Section 8 Expiring Contracts
Database, Dec.  1998. 

SIZE AND CONDITION OF THE
UNINSURED PORTFOLIO FOR 10 STATE
HOUSING FINANCE AGENCIES
========================================================== Appendix IV

This appendix augments chapter 4 by providing information on the size
of each state agency's uninsured portfolio, the physical and
financial condition of properties in the portfolio, and the number of
properties receiving special monitoring attention. 

Table IV.1 profiles the portfolios of the 10 state agencies we
reviewed. 

                               Table IV.1
                
                  Numbers of Properties, Project-Based
                 Section 8 Contracts, and Units in the
                  Portfolios of the 10 State Agencies
                            Reviewed by GAO

                                     Section 8       Total
State agency            Properties   contracts       units   Section 8
----------------------  ----------  ----------  ----------  ----------
California                     123         129       8,466       8,364
Illinois                       120         120      20,286      16,183
Maryland                        57          57       5,179       4,705
Massachusetts                  152         152      17,731      16,526
Michigan                       116         116      19,425      19,020
Minnesota                      232         232      15,028      13,353
New Hampshire                   63          63       2,268       2,244
Oregon                         122         122       4,045       4,010
Tennessee                       37          37       1,835       1,814
Wisconsin                      145         150      11,508      11,508
======================================================================
Total                        1,167       1,178     105,741      97,727
----------------------------------------------------------------------
Source:  GAO's analysis of information provided by the 10 state
agencies. 

Table IV.2 summarizes the physical condition ratings that the state
agencies assigned to properties in their portfolios.  Since the
agencies used their own criteria in developing and assigning the
ratings, it is difficult to compare the overall condition of their
respective portfolios. 

                                    Table IV.2
                     
                       Physical Condition of Properties, as
                          Rated by the 10 State Agencies

                     Superior      Satisfactory   Below average   Unsatisfactory
                  --------------  --------------  --------------  --------------
                  Proper          Proper          Proper          Proper
                    ties   Units    ties   Units    ties   Units    ties   Units
----------------  ------  ------  ------  ------  ------  ------  ------  ------
California             0       0     122   8,426       1      40       0       0
Illinois              89  13,583      31   6,703       0       0       0       0
Maryland              15   1,855      32   2,670       7     528       3      96
Massachusetts\a        0       0     146  17,466       6     265       0       0
Michigan              29   4,326      85  14,845       2     254       0       0
Minnesota             18   1,822     197  12,165      14     792       3     249
New Hampshire          4     200      59   2,068       0       0       0       0
Oregon\b              26     965      80   2,549      12     357       0       0
Tennessee             20   1,082      17     573       0       0       0       0
Wisconsin             43   3,872      97   7,393       4     237       1       6
================================================================================
Total                244  27,705     866  75,038      46   2,473       7     351
--------------------------------------------------------------------------------
\a The Massachusetts agency rates the physical condition of its
properties as meeting or not meeting the agency's standards. 
Massachusetts officials agreed to our classification of the physical
condition of these properties as ï¿½satisfactoryï¿½ or ï¿½below average.ï¿½

\b Information was not available for four of Oregon's properties. 

Source:  GAO's analysis of property condition ratings provided by the
10 state agencies. 

Table IV.3 summarizes the financial condition ratings assigned by 5
of the 10 housing agencies. 

                               Table IV.3
                
                  Financial Condition of Properties as
                      Rated by Five State Agencies

                                                                 Below
                                                              average,
                                                              poor, or
                               Superior or  Satisfactory  unsatisfacto
State agency                     excellent       or good            ry
----------------------------  ------------  ------------  ------------
California                               0           122             1
Michigan                               109             0             7
New Hampshire\a                         10            44             0
Tennessee                               15            22             0
Wisconsin\b                             40            94            10
======================================================================
Total                                  174           282            18
----------------------------------------------------------------------
\a New Hampshire did not provide ratings for nine properties. 

\b Wisconsin did not rate one of its properties. 

Source:  GAO's summary of information provided by the five state
agencies. 

Table IV.4 summarizes the overall condition of properties in the
portfolios of seven state agencies that rated the overall condition
of their properties.  The agencies typically considered such factors
as a property's physical, financial, and management condition in
assigning the ratings. 

                               Table IV.4
                
                Overall Condition of Properties as Rated
                        by Seven State Agencies

                                 Superior,                       Below
                                excellent,                  average or
                                  or above                unsatisfacto
State agency                       average  Satisfactory            ry
----------------------------  ------------  ------------  ------------
California                               0           122             1
Illinois                                89            29             2
Maryland                                15            30            12
Massachusetts\a                        137             0            13
Michigan                               109             0             7
New Hampshire                            9            54             0
Tennessee                               14            23             0
======================================================================
Total                                  373           258            35
----------------------------------------------------------------------
\a Massachusetts did not provide information on two properties. 

Source:  GAO's summary of information provided by the seven state
agencies. 

Table IV.5 summarizes the number of properties receiving special
monitoring attention by each state agency at the time of our inquiry. 

                               Table IV.5
                
                 Number of Properties Receiving Special
                 Monitoring Attention Compared With the
                 Total Number of Properties for Each of
                         the 10 State Agencies

                                                            Properties
                                 Properties in the   receiving special
State agency                    agency's portfolio          monitoring
------------------------------  ------------------  ------------------
California                                     123                   1
Illinois                                       120                   2
Maryland                                        57                   4
Massachusetts                                  152                   5
Michigan                                       116                   7
Minnesota                                      232                  11
New Hampshire                                   63                   2
Oregon                                         122                  12
Tennessee                                       37                   1
Wisconsin                                      145                   0
======================================================================
Total                                        1,167                  45
----------------------------------------------------------------------
Source:  GAO's analysis of information on special monitoring provided
by the 10 state agencies. 

RESULTS OF INSPECTIONS UNDER HUD'S
NEW INSPECTION PROGRAM <
=========================================================== Appendix V

Table V.1 summarizes the results of the physical inspections
completed under a new inspection program conducted by HUD's Real
Estate Assessment Center.  These inspections were completed by the
end of April 1999 at 4,513 of the properties in HUD's portfolio of
about 30,000 insured and uninsured properties. 

                               Table V.1
                
                Number of Properties Inspected by April
                   1999 and Their Physical Condition
                                Ratings

                                      Rating
                                ------------------
                        Number
                            of
                      properti
                            es
                      inspecte                                Excellen
Program                      d      Poor      Fair      Good         t
--------------------  --------  --------  --------  --------  --------
Uninsured:
Elderly and disabled     1,462        1%       10%       43%       46%
All other                  940        4%       19%       50%       27%
 uninsured\a
======================================================================
Subtotal                 2,401
Insured                  1,746        3%       16%       51%       30%
HUD-held                   366        6%       27%       47%       20%
======================================================================
Total                    4,513        3%       15%       48%       34%
----------------------------------------------------------------------
fna>The category ï¿½all other uninsuredï¿½ can include properties in the
state agency, new construction/substantial rehabilitation, rural
rental housing, loan management set-aside, property disposition, and
preservation programs.  HUD could not provide us with information on
the number of properties in each program included in this category. 

Source:  HUD's Real Estate Assessment Center. 

(See figure in printed edition.)Appendix VI
COMMENTS FROM THE DEPARTMENT OF
HOUSING AND URBAN DEVELOPMENT
=========================================================== Appendix V

(See figure in printed edition.)

(See figure in printed edition.)

(See figure in printed edition.)

(See figure in printed edition.)

(See figure in printed edition.)

The following are GAO's comments on the Department of Housing and
Urban Development's letter dated July 13, 1999. 

   GAO COMMENTS
--------------------------------------------------------- Appendix V:1

1.  The scope of the report reflects the requirements in Public Law
105-65 for a GAO study on the uninsured Section 8 portfolio.  The
mandate did not call for an evaluation of the Section 8 rental
assistance program.  It did, however, focus on the three topics
discussed in our report, including the information HUD has on the
Section 8 assistance provided to properties in the uninsured
portfolio and the physical and financial condition of these
properties.  To provide this information, we relied primarily on data
that were available from the Department. 

2.  See chapter 2, Agency Comments and Our Evaluation. 

3.  See chapter 2, Agency Comments and Our Evaluation. 

4.  The draft report contained information on the savings that bond
refundings have provided to the government and on how the state
agencies use their share of the savings.  Chapter 1 of the draft
report stated that as of September 30, 1998, the Department had
approved 245 refunding agreements that would provide $1.1 billion in
bond refunding savings over the lives of the Section 8 contracts.  Of
this amount, $633 million would be provided to the U.S.  Treasury,
and the remainder would be provided to state and local agencies. 
Furthermore, according to the draft report, the agencies use their
bond refunding savings to provide affordable housing for low-income
resident in their states.  This information also appears in the final
report. 

5.  See chapter 3, Agency Comments and Our Evaluation. 

6.  We do not agree with HUD that the report greatly exaggerates the
practical leverage the Department has to extract savings from state
and local housing agencies on transactions that HUD has no role in
approving or carrying out.  In fact, we believe that HUD does not
recognize the authority it does have to establish and enforce Section
8 program requirements.  The Department believes that since the state
agencies, unlike local agencies, do not need its approval to issue
tax-exempt bonds, then it cannot establish requirements for state
agencies issuing bonds for Section 8 properties.  We do not agree
that this limitation exists.  For example, at a minimum, the
Department could require the state agencies to notify the Department
when they plan to refund Section 8 bonds associated with Section 8
properties.  Such notification would provide the Department with the
information it needs to ensure compliance with shared savings
requirements.  Furthermore, the Department would not have to rely
solely on the state agencies to obtain this information.  State and
local bond activities are reported in national publications, such as
The Bond Buyer.\1 This publicly available information would allow the
Department to monitor the state agencies' compliance with a reporting
requirement, if necessary. 

7.  See comment 1 on the scope of this congressionally mandated
study.  We also note that HUD's assertions about the Section 8
program are not supported by objective data.  For example, as
discussed in the report, the portfolio reengineering program that the
Congress recently established is designed not only to reduce the
costs of expiring Section 8 contracts (reset them to market levels)
but also to address problems at financially and physically troubled
projects and to correct management and ownership deficiencies. 

8.  See chapter 3, Agency Comments and Our Evaluation. 

9.  See chapter 3, Agency Comments and Our Evaluation. 

10.  See chapter 3, Agency Comments and Our Evaluation. 

11.  The form and four worksheets cited by HUD did not include
examples of calculations showing how to adjust rents after bonds have
been refunded.  The form and worksheets were developed several years
before the Department introduced its rent adjustment policy. 

12.  See chapter 3, Agency Comments and Our Evaluation. 

13.  See chapter 3, Agency Comments and Our Evaluation. 

14.  This chronology of the dual fee issue, much of which appeared in
our draft report, does not alter the facts that 6 years have passed
since the Inspector General notified the Department of the dual fee
situation and HUD has not yet resolved this problem.  In addition,
HUD has neither identified all of the agencies that receive dual fees
nor taken steps to identify and enforce its dual fee regulation in
cases that are not covered by refunding agreements.  Therefore, we
did not revise the report in response to these comments. 

(See figure in printed edition.)Appendix VII

--------------------
\1 The Bond Buyer is a daily newspaper serving the municipal bond
industry. 

COMMENTS FROM THE NATIONAL COUNCIL
OF STATE HOUSING AGENCIES
=========================================================== Appendix V

(See figure in printed edition.)

(See figure in printed edition.)

(See figure in printed edition.)

The following are GAO's comments on the National Council of State
Housing Agencies' letter dated July 7, 1999. 

1.  See chapter 2, Agency Comments and Our Evaluation. 

2.  See chapter 3, Agency Comments and Our Evaluation. 

3.  See chapter 4, Agency Comments and Our Evaluation. 

4.  We agree that the Congress has placed a limit on certain Section
8 rent increases.  However, the National Council of State Housing
Agencies seems to have misidentified this requirement as a GAO
recommendation.  This report does not evaluate any potential effect
of the rent increase limit on housing finance agencies.  In this
report, we recommend that HUD (1) determine whether the state
agencies are required under the McKinney Act to share savings from
certain bond refundings, (2) provide the state and local agencies
with the appropriate methodology and sample calculations to follow
HUD's guidance on deducting the savings from refunding bonds before
calculating rent increases, and (3) enforce its prohibition of dual
fees unless there is a documented, sound, and equitable basis for
waiving the regulation. 

The National Council seems to imply that concerns about the financial
impact of the congressional limit on rent increases somehow justify
dual fees, even though the Section 8 program prohibits them. 
Concerns about the impact of this limit on properties' financial
viability could be grounds for requesting a waiver of the dual fee
prohibition but not for ignoring it.  Our recommendation that HUD
enforce its prohibition of dual fees unless there is a documented,
sound, and equitable basis for waiving the regulation would maintain
the integrity of the regulation while providing for exceptions in
cases of demonstrated need. 

GAO CONTACTS AND STAFF
ACKNOWLEDGEMENTS
======================================================== Appendix VIII

GAO CONTACTS

Stanley J.  Czerwinski, (202) 512-7631

Christine M.B.  Fishkin, (202) 512-6895

ACKNOWLEDGEMENTS

In addition to those named above, Mark H.  Egger, Elizabeth R. 
Eisenstadt, Diana Gilman, Richard A.  Hale, John T.  McGrail, Anthony
Radford, Joseph M.  Raple, Rose M.  Schuville, and Don Watson made
key contributions to this report. 

*** End of document. ***