Federal Power: Options For Selected Power Marketing Administrations' Role
in a Changing Electricity Industry (Chapter Report, 03/06/98,
GAO/RCED-98-43).

Pursuant to a congressional request, GAO reviewed various issues
concerning the role of certain power marketing administrations (PMA) and
other federal agencies in restructuring electricity markets, focusing
on: (1) whether the government operates them and the related electric
power assets in a businesslike manner; and (2) options that Congress and
other policymakers can pursue to address concerns about the PMAs' role
in restructuring markets and about their management.

GAO noted that: (1) although federal laws and regulations generally
require that the PMAs recover the full costs of building, operating, and
maintaining the federal power plants and transmission assets, in some
cases federal statues and Department of Energy's rules are ambiguous
about or prohibit the recovery of certain costs; (2) as GAO reported in
September 1997, for fiscal years 1992 through 1996, the federal
government incurred a net cost of $1.5 billion from its involvement in
the electricity-related activities of Southeastern, Southwestern, and
Western Area Power Administrations; (3) the $1.5 billion was the amount
by which the full costs of providing electric power exceeded the
revenues from the sale of power; (4) the availability of federal power
plants to generate electricity is below that of nonfederal plants
because the federal plants are aging and because the federal planning
and budgeting processes, as implemented, do not always ensure that funds
are available to make repairs when needed; (5) the resulting declines in
performance decrease the marketability of federal power; (6) to mitigate
these funding delays, the Bureau of Reclamation, Army Corps of
Engineers, PMAs and their preference customers have negotiated or are
negotiating agreements whereby customers pay for needed repairs in
advance; (7) the net cost to the Treasury and the decreased generating
availability of the federal power plants--when combined with the
competitive pressures on all electricity suppliers to decrease their
rates and the need to recoup some federal hydropower projects'
environmental costs--create varying degrees of risk that some of the
federal investment in certain hydropower plants and facilities will not
be repaid; (8) although the recovery of most of the federal investment
in Southeastern's, Southwestern's, and Western's hydropower-related
facilities is relatively secure, up to $1.4 billion out of about $7.2
billion of the federal investment in the electricity--related assets of
these PMAs is at some risk of nonrecovery; and (9) three general options
are available for the Bureau, the Corps, Southeastern, Southwestern, and
Western to address their roles in emerging restructured electricity
markets: (a) the Bureau and the Corps could continue generating and the
PMAs could continue marketing power as in the past; (b) the current
ownership structure could be maintained while improving how the federal
assets are managed and operated; and (c) the federal government could
divest the PMAs; the PMAs and the generating assets; or the PMAs, the
generating assets, and the dam reservoirs.

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

 REPORTNUM:  RCED-98-43
     TITLE:  Federal Power: Options For Selected Power Marketing 
             Administrations' Role in a Changing Electricity Industry
      DATE:  03/06/98
   SUBJECT:  Electric utilities
             Rural economic development
             Hydroelectric powerplants
             Property disposal
             Energy marketing
             Utility rates
             Electric power generation
             Financial management
             Facility repairs

             
******************************************************************
** 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 Requesters

March 1998

FEDERAL POWER - OPTIONS FOR
SELECTED POWER MARKETING
ADMINISTRATIONS' ROLE IN A
CHANGING ELECTRICITY INDUSTRY

GAO/RCED-98-43

Federal Role in a Changing Electricity Industry

(141044)


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

  CVP - Central Valley Project
  DOE - Department of Energy
  EIA - Energy Information Administration
  FERC - Federal Energy Regulatory Commission
  G&T - generation and transmission
  xID -
  IOU - investor-owned utility
  ISO - independent system operator
  kW - kilowatt
  kWh - kilowatthour
  MW - megawatt
  MWh - megawatthour
  NRC - Nuclear Regulatory Commission
  O&M - operations and maintenance
  PMA - power marketing administration
  POG - publicly owned generator
  REA - Rural Electrification Administration
  RUS - Rural Utilities Service
  TVA - Tennesee Valley Authority

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


B-278620

March 6, 1998

The Honorable Don Young
Chairman, Committee on Resources
House of Representatives

The Honorable John T.  Doolittle
Chairman, Subcommittee on Water and Power
Committee on Resources
House of Representatives

This report discusses various issues concerning the role of certain
power marketing administrations (PMA) and other federal agencies in
restructuring electricity markets.  We examined whether the
government operates them and the related electric power assets in a
businesslike manner and identified options that the Congress and
other policymakers can pursue to address concerns about the PMAs'
role in restructuring markets and about their management. 

As agreed with your offices, unless you publicly announce the
contents of this report earlier, we plan no further distribution of
this report until 30 days from the date of this letter.  At that
time, we will send copies of the report to other appropriate House
and Senate committees and subcommittees; interested Members of the
House and the Senate; the Administrators of the Southeastern,
Southwestern, and Western Area Power Administrations; the
Commissioner, Bureau of Reclamation; the Director for Civil Works,
U.S.  Army Corps of Engineers; and other interested parties.  We will
also make copies available to others upon request. 

If you or your staff have any questions, please call me at (202)
512-3841.  Major contributors to this report are listed in appendix
XI. 

Victor S.  Rezendes
Director, Energy, Resources, and
 Science Issues


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


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

From the early 1900s through September 30, 1996, the federal agencies
that generate and/or market electricity and that make or guarantee
loans to finance improvements to electricity systems incurred a debt
of about $84 billion.\1 \2 Like the other federal agencies, the
Southeastern, Southwestern, and Western Area power
administrations--responsible for $7 billion of this debt--face an
uncertain future as electricity markets restructure.  The Chairmen of
the House Committee on Resources and the Subcommittee on Water and
Power asked GAO to focus on these three power marketing
administrations (PMA) and to (1) examine whether the government
operates them and the related electric power assets in a businesslike
manner that recovers the federal government's capital investment in
those assets and the costs of operating and maintaining them and (2)
identify options that the Congress and other policymakers can pursue
to address concerns about the role of the three PMAs in emerging
restructured markets or to manage them in a more businesslike
fashion.  GAO's options also have implications for the Army's Corps
of Engineers (Corps) and the Department of the Interior's Bureau of
Reclamation (Bureau), which generate most of the power these PMAs
market.  As requested, the report also provides information about the
Tennessee Valley Authority (TVA), Rural Utilities Service, and
Bonneville Power Administration (Bonneville), which is contained in
appendixes I, II, and III, respectively. 


--------------------
\1 Dollars for the net costs are in constant 1996 dollars, unless
otherwise specified.  The $84 billion power-related debt is either
"direct" (owed directly to the Treasury--for example, the power
marketing administrations' appropriations that are repayable through
revenues earned from the sale of power) or "indirect" (owed to
nonfederal parties--for example, the Tennessee Valley
Administration's bonds that are held by nonfederal investors). 

\2 Federal Electricity Activities:  The Federal Government's Net Cost
and Potential for Future Losses (GAO/AIMD-97-110, Sept.  19, 1997). 


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

Traditionally, electric utilities were regulated monopolies;\3
however, they are now being subjected to competition as retail and
wholesale electricity markets restructure.  Under the traditional
compact between investor-owned utilities (IOU) and their state
regulators, IOUs were guaranteed monopolies within their service
areas.  In return, IOUs built generating and other facilities to
provide electricity to all existing and future customers in their
service areas.  Under this traditional regulation, the states
approved electricity rates that reflected the utilities' costs of
building and operating their facilities and that included approved
financial returns on these investments.  Competition was introduced
into wholesale electricity markets by the Public Utilities Regulatory
Policies Act of 1978, which allowed entities that were not utilities
to compete with utilities.  Many of these nonutilities generate power
using relatively inexpensive natural-gas-fired generating
technologies.  Subsequently, the Energy Policy Act of 1992 called for
utilities to transmit power generated by outside entities to
wholesale customers inside of their service areas, thus introducing
competition.  Retail electricity markets are also restructuring; at
least 17 states are implementing measures that would allow customers
to choose their electricity suppliers.  According to the Department
of Energy (DOE), by 2015, competition will cause retail electricity
rates to drop by 6 percent to 19 percent below the level they would
have been in the absence of competition. 

The federal government began to market electricity after the Congress
authorized the construction of dams and established major water
projects, primarily in the 1930s to the 1960s.  The Bureau and the
Corps operate these projects to provide or manage water for such
multiple purposes as irrigation, flood control, navigation,
recreation, water supply, and environmental enhancement.  These
agencies also generate electricity at about 130 hydropower plants
located at federal water projects.  The PMAs\4 sell the power that is
not used for projects' purposes\5 to "preference
customers"--cooperatives and public bodies, such as municipal
utilities, irrigation districts, and military installations. 
Historically, one of the important reasons for selling this power was
to electrify portions of rural America that IOUs were reluctant to
serve because of cost considerations.  Rural America is now
electrified.  The federal government today markets about 10 percent
of the nation's power through the PMAs as well as TVA--a wholly owned
federal corporation that generates and markets federal power in
Tennessee and parts of six other southeastern states. 

The power the PMAs sell is relatively inexpensive.  In 1990 through
1995, Southeastern's, Southwestern's, and Western's average revenues
per kilowatthour (kWh)\6 were about 40 percent less than the other
power providers' average revenues.  The PMAs' rates are generally to
be set at the lowest levels practicable, consistent with sound
business principles, while generally still recovering the costs of
producing, transmitting, and marketing power, including the
repayment, with interest, of the federal investment in the power
generating facilities and other debt.  However, under current federal
laws, an applicable DOE order, and repayment practices, certain costs
are excluded from the PMAs' rates, such as the full costs of (1)
interest to finance the power facilities; (2) pension and
postretirement benefits for the PMAs', the Bureau's, and the Corps'
employees; and (3) the construction of a few federal power projects. 
Some costs to mitigate the environmental damages caused by certain
federal water projects, including their hydropower plants, also must
be excluded.\7


--------------------
\3 Electric utilities function as monopolies and provide electricity
to customers in their exclusive service areas.  Three types of
electric utilities exist:  (1) investor-owned utilities, which
constitute only about 8 percent of nation's 3,200 electric utilities
but have over three quarters of the sales to ultimate customers; (2)
932 customer-owned rural electric cooperatives; and (3) 2,014
publicly owned utilities.  In addition, nonutilities (or nonutility
generators) exist that have no designated service areas and generate
power which they sell in wholesale markets.  They may generate power
primarily for their own use (e.g., at petroleum refineries) and sell
the excess power, or generate power primarily to sell it. 

\4 In addition to Southeastern, Southwestern, and Western, Bonneville
operates in the Pacific Northwest and is the oldest and largest PMA. 
The Alaska Power Administration (Alaska) is the smallest PMA.  Unlike
the other PMAs, Alaska generates its own electricity.  The Congress
passed a law in 1995 authorizing the divestiture of Alaska's power
assets; the divestiture is ongoing. 

\5 For example, for pumping water to fields being irrigated. 

\6 A watt is the basic unit used to measure electric power.  A
watthour is equal to a watt of power applied for 1 hour.  A
kilowatthour (kWh) is 1,000 watthours. 

\7 See GAO/AIMD-97-110.  For example, Western incurred costs of $53.8
million in fiscal years 1992 through 1996 to buy power for its
customers because the Shasta project in California released water to
protect fisheries.  However, the 1991 Energy and Water Development
Appropriations Act specified that these costs not be allocated for
repayment through PMA customers' electric rates. 


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

Although federal laws and regulations generally require that the PMAs
recover the full costs of building, operating, and maintaining the
federal power plants and transmission assets, in some cases federal
statutes and DOE's rules are ambiguous about or prohibit the recovery
of certain costs.  As GAO reported in September 1997, for fiscal
years 1992 through 1996, the federal government incurred a "net cost"
of $1.5 billion from its involvement in the electricity-related
activities of Southeastern, Southwestern, and Western.  The $1.5
billion was the amount by which the full costs of providing electric
power exceeded the revenues from the sale of power.  In addition, the
availability of federal power plants to generate electricity is below
that of nonfederal plants because the federal plants are aging and
because the federal planning and budgeting processes, as implemented
by the Bureau and the Corps, do not always ensure that funds are
available to make repairs when needed.\8 The resulting declines in
performance decrease the marketability of federal power.  To mitigate
these funding delays, the Bureau, the Corps, the PMAs, and their
preference customers have negotiated or are negotiating agreements
whereby customers pay for needed repairs in advance.  The net cost to
the Treasury and the decreased generating availability of the federal
power plants--when combined with the competitive pressures on all
electricity suppliers to decrease their rates and the need to recoup
some federal hydropower projects' environmental costs--create varying
degrees of risk that some of the federal investment in certain
hydropower plants and facilities will not be repaid.  For example,
although the recovery of most of the federal investment in
Southeastern's, Southwestern's, and Western's hydropower-related
facilities is relatively secure, up to $1.4 billion out of about $7.2
billion of the federal investment in the electricity-related assets
of these PMAs is at some risk of nonrecovery.  For example, at the
Corps' power plants at the Truman project in Missouri and the Russell
project in South Carolina, over $500 million of the federal
investment to build these assets currently is not being recovered
through the power rates charged by Southwestern and Southeastern,
respectively.  Under the PMAs' existing rate-setting practices, these
costs cannot be placed into the rates until the pumpback units at the
Russell project and the turbines at the Truman project come into
service as designed.  Because operation of these power plants would
kill large numbers of fish, the affected units cannot be placed into
service as intended until this issue is resolved.  According to the
Corps, repairs to these two projects are to be completed by the end
of fiscal year 1999. 

Three general options are available for the Bureau, the Corps,
Southeastern, Southwestern, and Western to address their roles in
emerging restructured electricity markets.  First, the Bureau and the
Corps could continue generating and the PMAs could continue marketing
power as in the past.  This option perpetuates the net costs to the
government and does not decrease the risk that the federal investment
in certain of the government's electricity-related assets will not be
fully recovered.  Nor does it resolve questions about the continued
role of federal power in restructuring markets, such as why the
government continues to provide power to rural areas that are already
electrified and why it sells this low-cost power only to customers in
the South and West.  This option continues to balance the existing
multiple uses of water and allows time for policymakers to consider
changes that can be made to the operations of the Bureau, the Corps,
and the PMAs. 

Second, the current ownership structure could be maintained while
improving how the federal assets are managed and operated, including
making changes to better recover the operations and maintenance costs
as well as the federal investment in the power assets.  This option
has many suboptions, such as revising the federal agencies' planning
and budgeting processes to improve the timeliness and certainty of
funding for repairs; modifying the PMAs' rate-setting and repayment
methodologies to better recover costs; restructuring the hydropower
program, perhaps in the form of federal corporations, to improve its
efficiency; and freeing the federal agencies from certain legal and
administrative requirements.  Drawbacks include not resolving the
concerns about the role of federal power in restructuring markets. 

Third, the federal government could divest the PMAs; the PMAs and the
generating assets; or the PMAs, the generating assets, and the dams
and reservoirs.  Any of these actions would end the government's role
in selling power in a competitive market.  Depending on the sale's
terms and conditions and the price obtained, a divestiture may or may
not recover the government's investment in hydropower-related assets. 
Divestiture is complex because steps would be needed to balance the
multiple purposes of the water projects and to accommodate related
interests.  Also, the effect of a divestiture on the PMAs' customers'
rates would need to be considered.  Finally, some divestitures could
result in sales proceeds that do not recover the federal investment. 
For example, if the government transferred some liabilities, imposed
restrictions after a sale, or limited the availability of water to
generate electricity, a lower price for the assets could result. 


--------------------
\8 See, for example, Federal Power:  Outages Reduce the Reliability
of Hydroelectric Power Plants in the Southeast (GAO/T-RCED-96-180,
July 25, 1996). 


   GAO'S ANALYSIS
---------------------------------------------------------- Chapter 0:4


      SOUTHEASTERN'S,
      SOUTHWESTERN'S, AND
      WESTERN'S POWER PROGRAMS
      OPERATE AT A NET COST, HAVE
      GENERATING ASSETS THAT NEED
      REPAIR, AND POSE SOME RISK
      THAT THE FEDERAL INVESTMENT
      MAY NOT BE REPAID
-------------------------------------------------------- Chapter 0:4.1

As GAO recently reported, the federal program to generate and market
power and to make or guarantee loans to rural utilities operates at a
net cost of billions of dollars to the Treasury.  For Southeastern,
Southwestern, and Western, this net cost totaled about $1.5 billion
in fiscal years 1992 through 1996 because these PMAs' power rates do
not recover all of the costs associated with the production,
transmission, and sale of power.  It is important to note that the
three PMAs were generally following applicable laws and regulations
applying to the recovery of costs; however, in some cases, federal
statutes and an applicable DOE order are ambiguous about or prohibit
the recovery of certain costs.  To mitigate the funding delays that
characterize the agencies' planning and budgeting processes, the
Bureau, the Corps, and the PMAs have instituted efforts to collect
funding from preference customers to pay for needed repairs of the
federal hydropower assets in a more timely and predictable fashion. 
For example, Western's preference customers have agreed to finance
repairs at the Bureau's Shasta plant in California by depositing up
to $21 million in an escrow account to pay for the work.  According
to Bureau officials, customers who contributed funds will be issued
credits on their monthly power bills from Western, while those who
did not contribute will not be issued these credits. 

This $1.5 billion of net costs included net financing costs of about
$1.2 billion.  These net financing costs occurred mostly because (1)
much of Southeastern's, Southwestern's, and Western's outstanding
appropriated debt\9 was provided at low interest rates while the
Treasury's financing costs for this money were higher and (2) these
PMAs, under an applicable DOE order, generally repay debt with higher
interest rates before repaying lower-rate debt from the Treasury,
which causes the Treasury to incur additional, higher costs.  Before
1983, the PMAs generally incurred appropriated debt at below-market
rates.  The average interest rate on the PMAs' outstanding
appropriated debt (about 3.5 percent) is substantially below the
average rate the Treasury has incurred (about 9 percent\10 ) to fund
federal programs.  In addition, Southeastern's, Southwestern's, and
Western's rates did not recoup about $82 million of the cost of
providing retirement benefits to their and the operating agencies'
employees and about $138 million in interest related to power
generating projects that are incomplete, are under construction, or
were canceled.  A balance of about $157 million for other costs in a
variety of categories was also not recouped. 

The Bureau's and the Corps' power plants have become less available
to generate electricity than those of other utilities,\11 which makes
the PMAs' power less attractive to customers at a time when
competition is giving them more opportunities to buy reasonably
priced power from a variety of suppliers.  Although power plants'
maintenance needs differ by location, within the operating agencies
federal power plants go off line for two basic reasons.  First, the
age of the plants (the Bureau's plants average about 50 years in
service and the Corps' about 30 years) increases the need for
repairs.  Second, the federal planning and budgeting processes, as
implemented by the Bureau and the Corps, do not always provide
funding to repair the federal power assets when it is needed,
delaying some repairs and also causing the power plants to become
less available to provide power.  Specifically, the Bureau's and the
Corps' field locations identify improvements for, estimate the costs
of, and develop the budget proposals for not only hydropower
facilities but also other facilities, such as dams, irrigation
systems, and recreational facilities.  Given these competing
purposes, repairs of hydropower facilities sometimes take lower
priority than other items.  Also, budget requests to fund hydropower
repairs have been cut by 10 percent to 15 percent to reduce the
federal deficit.  In GAO's view, maintaining this power's
availability is needed to ensure that the power revenues recover as
much of the federal costs and investment as possible.  Moreover, if
the Congress and other policymakers decide to divest the federal
power assets, then maintaining the power's availability could
facilitate the divestiture; however, the government would not want to
spend so much on repairing and upgrading its assets that the amount
spent exceeded any increases in the sales proceeds or the value of
those improvements. 

The large, recurring net costs to the Treasury of operating the
federal hydropower program, along with the decreased availability of
the generating assets, contribute to the risk that the taxpayers'
investment in the federal hydropower assets will not be recovered. 
Other factors, too, increase the risk of nonrecovery.  One general
factor is the onset of market competition, which is holding down
market rates.  At the same time, the PMAs' electricity rates at some
projects face increased costs; these include (1) the costs of
mitigating the damages to fish and wildlife habitat caused by
generating hydropower and (2) purchasing power to sell to the PMAs'
power customers when, to protect the environment, federal power
plants reduce the electricity generated.  In general, at
Southeastern, Southwestern, and Western, most of the federal
investment is relatively secure.  Because these PMAs sell power at
low rates, it is relatively easy to sell, and the resulting revenues
facilitate the recovery of the federal investment.  However, as GAO
recently reported, up to about $1.4 billion of the investment in the
hydropower-related assets of these PMAs (out of a total federal
investment of about $7.2 billion in their power assets) is at some
risk of nonrecovery.  In addition to the previous examples of the
Truman and Russell projects, for which over $500 million may not be
recovered, about $464 million that the Bureau invested in power
generating capacity and water storage within the Pick-Sloan Missouri
Basin Program\12 may not be recovered without congressional action. 
These assets were designed to serve future irrigation projects, but
under existing legislation, about $464 million cannot be recovered
through Western's electricity rates until the projects come into
service.  However, according to the Bureau, these projects are
infeasible and likely will never come into service. 


--------------------
\9 GAO uses the term "appropriated debt" because the PMAs and TVA are
required to repay appropriations used for capital investments, with
interest.  However, the Department of the Treasury does not
technically consider these reimbursable appropriations to be lending. 

\10 This rate is the weighted average interest rate on the Treasury's
entire outstanding bond portfolio (10- to 30-year maturities) as of
September 30, 1996.  GAO used this interest rate because it reflects
the Treasury's average interest rate on outstanding long-term debt
and because this debt most closely matches the terms of the PMAs'
appropriated debt. 

\11 According to data provided by the Corps, its hydropower plants
were available to provide power 92.9 percent of the time in fiscal
year 1987 but only 87.9 percent of the time in fiscal year 1995. 
However, the availability of these plants improved to 88.4 percent in
fiscal year 1996 and 89 percent in fiscal year 1997.  The Corps
attributes this improvement, in part, to $450 million committed to
repair its hydropower assets from fiscal year 1993 through fiscal
year 2007.  The Bureau's plants were available only 83.4 percent of
the time in 1994, compared with the industry's average of 89 percent. 
According to Bureau officials, the availability of the Bureau's
plants over the last 3 years has improved over the average
availability of the last 15 years. 

\12 The Program consists of 13 of the Corps' and the Bureau's
hydropower plants and associated irrigation projects, among other
assets, located in the northern basin of the Missouri River.  Western
sets rates that are designed to recover not only the federal capital
investment in the power system, but also part of the federal
investment in irrigation, as well as other costs. 


      OPTIONS EXIST TO ADDRESS THE
      FEDERAL ROLE IN A MORE
      COMPETITIVE MARKET
-------------------------------------------------------- Chapter 0:4.2

Three general options exist to address the federal role in
restructuring markets:  (1) maintaining the status quo of federal
ownership and operation of the power generating projects, (2)
maintaining the federal ownership of these assets but improving how
they are operated, and (3) divesting these assets. 


         MAINTAINING THE STATUS
         QUO
------------------------------------------------------ Chapter 0:4.2.1

Maintaining the status quo perpetuates the recurring net costs to the
Treasury and the risk that some of the federal investment will not be
repaid.  In addition, this option does not resolve concerns about the
continued role of federal power in restructuring electricity markets. 
Specifically, the government's power program has successfully
electrified rural areas; therefore, an original justification for the
government to provide power in these areas has passed.  Moreover, one
could question the equity of the PMAs' providing low-cost power to
customers in 34 states primarily in the South and West but not to
other areas.  IOUs and other critics of the PMAs have also argued
that, as federal agencies, the PMAs have advantages that the IOUs do
not have.  As GAO's work has shown, the PMAs have charged rates that
do not recover all of the government's costs of generating,
transmitting, and marketing power.  Also, as federal agencies, the
PMAs do not pay income taxes, are not overseen by state regulators,
and have more flexibility to set rates than nonfederal utilities. 

The status quo continues the federal role in balancing the multiple
uses of water and allows policymakers time to study these issues
before they change the operations and/or ownership of the water
projects and power assets.  How water is used affects wide geographic
areas across state lines and has a significant impact on people's
lives.  It affects such things as how much water will be available to
accommodate the expansion of metropolitan areas, how much water will
be used to protect endangered species, and how much water will be
needed to protect shellfishing--in Apalachicola Bay, Florida, for
instance.  The Bureau and the Corps generate power while balancing
these impacts.  Any decisions that federal policymakers reach about
changing how power is generated or how the water projects are managed
or owned will need to consider the impacts on the uses of the water
and the beneficiaries of the projects. 


         IMPROVING THE MANAGEMENT
         OF THE POWER PROGRAM
         WITHIN FEDERAL OWNERSHIP
------------------------------------------------------ Chapter 0:4.2.2

Under the second option, the management of the federal power assets
could be improved while they remain under federal ownership. 
Properly implemented, such improvements could help promote the
recovery of the operation and maintenance costs of the power program
as well as the federal investment in the power assets.  It could also
help prepare these assets for divestiture if the Congress decides to
divest them.  However, this option does not address the questions,
previously discussed, about the federal government's participation in
a commercial activity.  Depending on how they are structured, some
reforms may decrease opportunities for oversight by the Congress. 

This option includes several suboptions.  First, the Bureau's and the
Corps' planning and budgeting processes could be revised to secure
funding more quickly and predictably than is currently the case to
repair the hydropower assets.  The budgeting process is lengthy and,
as described, has required cuts of 10 percent to 15 percent of the
agencies' budget requests.  Consequently, funding for repairs is
uncertain and sometimes is not available when needed.  One solution
would be to institute revolving funds for the PMAs.  Under this
arrangement, a one-time permanent appropriation is replenished
through revenues that are earned by selling power or other services
and credited directly to a fund, instead of being replenished through
annual appropriations.  These funds, which could be used to pay for
operations, maintenance, repairs, and replacements for the power
plants and other assets, enable funding to occur that is not subject
to the uncertainties of the operating agencies' budget processes. 
Funding for needed repairs is approved faster and is made available
with more certainty, according to agency and PMA customer association
officials.  Several water projects that generate power now have
revolving funds, which the Congress could extend to other projects. 
Also, under agreements with the agencies, the PMAs' customers can
provide up-front funding for capital repairs and improvements.  For
example, the Bureau and Western have negotiated or are negotiating
such arrangements at several water and hydropower projects involving
tens of millions of dollars of funding for repairs. 

Second, the Congress or the Secretary of Energy could change how the
PMAs' revenue requirements and rates are established to more fully
recover the costs of generating, transmitting, and marketing power. 
Where prudent, the Congress or the Secretary of Energy could direct
or authorize the PMAs to charge higher rates to enable them to better
recover costs and reduce the risk that the federal investment will
not be repaid.\13 In fiscal year 1998, Southeastern, Southwestern,
and Western are to take a step in this direction by beginning the
process of recovering, through their rates, the full costs of the
pension and postretirement health benefits of their employees.  The
Congress or the Secretary of Energy could also direct DOE to revise
the methodology for the PMAs' repayment of their debt, thereby
increasing the PMAs' electric rates, power revenues, the amount
repaid to the Treasury, and the rate of repayment to the Treasury. 
Because the amount of hydropower generated can vary from year to
year, federal laws and an applicable DOE order allow the PMAs to
defer repayment of the annual expenses during some "low water
years."\14 The PMAs also generally repay their highest
interest-bearing debt first rather than the older lower-rate debt
from the Treasury.  Consequently, their electricity rates are lower
than otherwise, with the older debt deferred.  The repayment of the
federal investment is also lower.  This situation results in
additional costs to the Treasury because interest rates on the
outstanding federal investment are substantially below the interest
rates the Treasury incurs to provide funding to the PMAs and other
federal programs.  Repaying the federal investment faster could
decrease the Treasury's interest costs and could decrease the amount
of investment at risk of nonrecovery.  However, policymakers may need
to consider the impact of any rate increases on the PMAs' customers. 

Third, the Congress could also restructure the PMAs as federally
owned corporations.  With this action, the PMAs could finance repairs
and improvements more expeditiously and predictably than under the
federal budget process because the PMAs would self-finance and would
require fewer external approvals and oversight.  Establishing a
government corporation could also serve as an interim step toward
divesting the federal hydropower assets. 

Finally, the PMAs, the Bureau, and the Corps could be exempted from
certain legal and administrative requirements that, according to
agency officials, cause them to operate inefficiently and can cause
the PMAs' power rates to be higher than otherwise.  According to a
May 1996 study by Western, if the Congress had authorized Western to
pay prevailing local wages for its service contracts in fiscal years
1992 through 1995, instead of the higher wages prescribed by law, it
could have saved about $6.2 million per year. 


--------------------
\13 It should be noted, however, that along with such factors as
costs incurred to mitigate environmental impacts, these changes could
place upward pressure on rates for some rate-setting systems to the
point where they exceed regional rates.  In a competitive market, any
measure that increased the PMAs' rates would jeopardize the PMAs'
ability to sell power and repay the federal investment. 

\14 The amount of hydropower generated varies from year to year,
given changes in water flows.  Deferred amounts bear a current
interest rate and are to be repaid on a priority basis before all
other investment.  Repayment is to be accelerated during good water
years. 


         DIVESTING THE PMAS AND
         HYDROPOWER ASSETS
------------------------------------------------------ Chapter 0:4.2.3

Under the third option, divesting the PMAs and federal power assets
would eliminate the government's presence in a commercial activity
and, depending on a divestiture's terms and conditions and the price
obtained, could produce both a net gain and a future stream of tax
payments to the Treasury.  The Congressional Budget Office recently
estimated that a sale of Southeastern, Southwestern, and Western and
the related hydropower assets would result in revenues of between $8
billion and $11 billion; these revenues might not be enough to
recover the government's investment in hydropower-related assets.\15
Divestitures of government assets have been accomplished recently in
the United States and also overseas; GAO's March 1997 report
concluded that divesting the federal hydropower assets would be
complicated but not impossible.\16 Such a transaction would need to
balance the multiple purposes of the water project as well as other
claims on the water.  The federal responsibility for balancing water
use among the authorized purposes and other public policy goals would
not necessarily end after a divestiture.  Depending on the
divestiture's conditions, balancing a project's purposes or
accommodating other public considerations may affect a project's
operation afterwards and thereby lead to continued liability for
taxpayers. 

Some of Southeastern's, Southwestern's, and Western's customers are
concerned that a sale would significantly raise their rates--the
PMAs' average revenues of under 2 cents per kWh were at least 40
percent less than the average revenues for nonfederal utilities in
1990 through 1995.  Therefore, how a divestiture could affect
preference customers' rates needs to be considered.  In general,
because most preference customers buy only a small portion of their
total power from these PMAs, GAO estimates that most of them would
experience relatively small changes in their wholesale rates.  For
example, if, after a divestiture, the rates for the PMAs' power
increase to market rates, about two-thirds of these PMAs' preference
customers would experience rate increases of 25 percent (roughly 0.5
cents per kWh) or less.  If these preference customers passed their
rate increases directly on to the end-users they serve, their average
residential customers would experience increases in their electricity
bills of no more than $4.17 per month. 

However, some preference customers--in particular ones that purchase
most of their power from the PMAs--could experience much larger
increases.  For example, in 1995, 35 percent of Western's preference
customers purchased more than half of their electricity from the PMA. 
Correspondingly, GAO estimates that about one-fifth of Western's
customers may see their rates increase by more than 75 percent. 
Similarly, about 27 percent of Western's preference customers would
see rate increases exceeding 1.5 cents per kWh.  However, although
some preference customers could initially experience significant rate
increases, the government could mitigate these increases through such
mechanisms as rate caps.  It should also be noted that, after a
divestiture, preference customers would pay the same market rates as
neighboring utilities who lack access to PMA power. 

A divestiture's goals would affect how the government proceeds in
divesting its hydropower assets.  In addition, trade-offs and the
terms and conditions of any divestiture would need to be considered
carefully so as not to jeopardize the government's finances.  If the
government decided to obtain a larger price for its assets, it could
choose to retain many of the liabilities and related costs--for
example, by retaining the costs of mitigating environmental damages. 
In contrast, if the government transferred these liabilities and
costs, the prices obtained for its assets would likely be less than
if it kept these liabilities and costs. 


--------------------
\15 Should the Federal Government Sell Electricity?,Congressional
Budget Office, (Nov.  1997.)

\16 Federal Power:  Issues Related to the Divestiture of Federal
Hydropower Resources (GAO/RCED-97-48, Mar.  31, 1997). 


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

This report contains no recommendations. 


   AGENCY COMMENTS
---------------------------------------------------------- Chapter 0:6

GAO provided a draft of this report to DOE (which represented the
views of Southeastern, Southwestern, and Western), the Department of
the Interior (including the Bureau), the Department of Defense
(including the Corps), Bonneville, and the Federal Energy Regulatory
Commission.  The comments of DOE, Interior, the Corps, Bonneville,
and the Federal Energy Regulatory Commission, and GAO's responses to
those comments are included in appendixes VI, VII, VIII, IX, and X,
respectively. 

In commenting on the report, DOE concurred that in some cases the
PMAs do not recover the full costs of marketing federal power as
defined by GAO.  However, according to DOE, GAO overstates these
costs because it overstates the amount of investment that was
financed with interest rates that were less than the Treasury's cost
of borrowing.  GAO does not agree that it overstates these costs. 
The interest rate used by the PMAs in calculating the amounts to be
repaid through their power rates was less than the Treasury's cost of
borrowing those funds.  Furthermore, GAO believes that by not
limiting the estimate of the financing costs to differences in
interest rates, GAO's methodology accurately captures the full amount
of the financing costs.  DOE concurred that some portion of $1.4
billion of federal investment in power-related assets is at risk of
not being repaid through PMAs' power rates.  However, DOE believes
that GAO overstates the amount of investment at risk of not being
repaid.  GAO believes its assessment of risk is accurate and did not
change its assessments because DOE did not provide information that
would allow GAO to change its assessments.  More detailed responses
to DOE's comments are found in appendix VI.  DOE also provided
general policy comments and technical clarifications that are
incorporated in the report as appropriate. 

Interior provided general and specific comments on the report.  The
most significant comment was that the reduced percentage of time the
Bureau's plants could generate power was not an indication of
inadequate maintenance, but rather caused by the need, under
statutes, to manage water projects to satisfy multiple uses, such as
irrigation.  GAO disagrees that the need to manage water projects for
multiple uses necessarily leads to a reduced percentage of time to
generate power.  GAO notes that the percentage of time a plant can
generate electricity is not affected by nonpower uses of water but
rather by scheduled and unscheduled repairs of the plants. 

Defense presented detailed, technical, and clarifying comments that
GAO incorporated into the report as appropriate.  For example,
Defense provided GAO with recent data regarding the improved
performance of the Corps' hydropower plants.  It also provided
information about $450 million in funding to repair and rehabilitate
those plants.  GAO incorporated this information into the report. 

In commenting on the report, Bonneville stated, most significantly,
that its activities do not impose substantial net costs to the
federal government.  GAO disagrees because Bonneville's operations
entailed net financing costs to the government of about $377 million
in fiscal year 1996. 

The Federal Energy Regulatory Commission provided GAO with technical
comments regarding the implementation of the Commission's Order 888
and its applicability to the PMAs.  GAO incorporated those comments
into the report. 


INTRODUCTION
============================================================ Chapter 1

The electricity industry has been predominantly monopolistic and
noncompetitive.  Utilities (primarily investor-owned utilities--IOU)
build power plants and power lines to provide all of the electricity
needed by all existing and future customers in their exclusive
service areas.  Regulators in the states allow utilities to charge
electricity rates that give them a regulated, specified level of
return on these investments. 

IOUs were initially reluctant to provide electricity to rural areas,
mostly because the sparse population made it difficult for them to
recover their costs and to earn a profit.  The federal government has
played an important role in the traditional market by selling power
to rural America.  The Department of the Interior's Bureau of
Reclamation (the Bureau) and the Department of the Army's Corps of
Engineers (the Corps) generate electricity at hydropower plants
located at major federal water projects.  The Department of Energy's
(DOE) power marketing administrations (PMA) generally sell this power
in wholesale markets, mostly to publicly and cooperatively owned
utilities that, in turn, sell power to end-use (retail) consumers. 
The PMAs repay the federal investment in the government's power
plants, power lines, and related assets through the revenues they
earn by selling power.  The Tennessee Valley Authority (TVA), a
federal corporation, generates and markets power throughout Tennessee
and parts of six other southeastern states.  Moreover, the Department
of Agriculture's Rural Utilities Service (RUS) makes and guarantees
loans to rural utilities to finance the construction and development
of electric power systems.  Although critics question the federal
government's role in providing power or in financing improvements to
rural utility systems as markets restructure, the activities
continue. 

However, the traditional structure of the electricity industry has
begun to change.  Legislation and new generating technologies have
introduced increased competition into the market, changing the
environment in which the PMAs must operate successfully if they are
to repay the federal investment in the power program. 


   STRUCTURE OF THE ELECTRIC POWER
   INDUSTRY
---------------------------------------------------------- Chapter 1:1

Federal and state agencies regulate the activities of electric
utilities.  Traditionally, electricity service was viewed as a
"natural monopoly":  A central source of power was seen as the most
efficient way of generating, transmitting, and distributing
electricity at a reasonable cost.  Under the traditional regulatory
compact between electric utilities and their state regulators,
electric utilities were guaranteed monopolies within their exclusive
service areas and regulated rates of return on their capital
investments.  In return, these utilities built generating and other
facilities to provide all of the electricity needed by all current
and future customers in their service areas.  Under traditional
"cost-of-service" regulation, electricity rates approved by state
regulators reflected the utilities' costs of building new generating
plants and operating the power system.  As shown in table 1.1, IOUs
dominate the electricity markets:  Although they account for only
about 8 percent of the nation's almost 3,200 electric utilities, they
have over 75 percent of utility sales to ultimate customers and over
77 percent of total utility power generation.  Most IOUs sell power
at retail rates to several different classes of consumers and at
wholesale rates to other utilities, including other IOUs; federal,
state, and local government utilities; public utility districts; and
rural electric cooperatives. 

The traditional regulatory role of the federal and state governments
was established under the Constitution and developed by federal law. 
Specifically, the Federal Power Act (formerly the Federal Water Power
Act), which was enacted in 1920, and the Public Utility Holding
Company Act established a regime of regulating electric utilities
that gave specific and separate powers to the states and the federal
government.  State regulatory commissions (generally called "public
utility" or "public service commissions") regulate utilities'
activities within state boundaries, including the setting of
wholesale and retail electric rates.  At the federal level, the
Securities and Exchange Commission regulates interstate electric
utility holding companies by requiring them to register and divest
holdings so that each company becomes a single consolidated system
serving a specific geographic area.  In addition, the Commission
regulates how the holding companies issue and acquire securities. 
Under the Federal Power Act, the Federal Energy Regulatory Commission
(FERC), formerly the Federal Power Commission, regulates interstate
aspects of the electric utility industry, including financial
transactions, wholesale rates, and interconnection and transmission
arrangements. 

In addition to IOUs, 932 customer-owned rural electric cooperatives
and 2,014 publicly owned utilities provided power in 1996.  Most
rural electric cooperatives, usually formed and owned by residents of
rural areas, distribute electricity only to their members.  Operating
throughout the nation except for Connecticut, Hawaii, and Rhode
Island, cooperatives constituted 29 percent of all the nation's
electric utilities in 1996.  Publicly owned electric utilities are
nonprofit state and local government agencies, such as municipal
utilities, state authorities, public power districts, and irrigation
districts.  DOE views publicly owned power as providing competition
for IOUs and as charging power rates against which the power rates of
IOUs can be compared.  In 1996, almost 63 percent of all electric
utilities in the nation were publicly owned utilities.  Cooperatives
and publicly owned utilities buy power from wholesale providers for
sale to retail customers.  However, some cooperatives and publicly
owned utilities also generate their own power and transmit it to
other utilities or distribute it to their own retail customers.  The
generation and share of the national energy supply for these types of
utilities are provided in table 1.1. 



                               Table 1.1
                
                Number of Electric Utilities by Class of
                           Ownership in 1996

                            Percen              Percen          Percen
                              t of       Net\a    t of  Sales\    t of
Type of utility     Number   total  generation   total       a   total
------------------  ------  ------  ----------  ------  ------  ------
Investor-owned         243     7.6     2,374.4    77.2  2,346.    75.7
                                                             1
Cooperatives           932    29.1       139.2     4.5   258.4     8.3
Publicly owned       2,014    63.0       266.1     8.6   450.9    14.5
Federal               10\b     0.3       297.9     9.7    45.6     1.5
======================================================================
Total                3,199   100.0     3,077.4   100.0  3,101.   100.0
                                                             1
----------------------------------------------------------------------
\a Net generation and sales are in millions of megawatthours (MWh). 
One MWh equals 1,000 kilowatthours (kWh).  One kWh equals 1,000
watthours.  One watthour equals the total amount of electricity used
in 1 hour by a device that uses one watt of power for continuous
operation.  A watt is the basic unit used to measure electric power. 

\b In addition to the five PMAs, the Bureau, the Corps, and TVA,
DOE's Energy Information Administration (EIA) classifies the
Department of the Interior's Bureau of Indian Affairs and the
International Water and Boundary Commission as federal electric
utilities. 

Source:  Developed by GAO from data provided by EIA. 


      THE ROLE OF THE FEDERAL
      GOVERNMENT IN TRADITIONAL
      ELECTRICITY MARKETS
-------------------------------------------------------- Chapter 1:1.1

The federal government has played a significant role in the
development of electricity markets.  Because it was too expensive for
IOUs to serve rural areas, federal power agencies provided power to
those areas.  In addition, the government provided financing to rural
utilities to assist them in building and maintaining electricity
distribution systems that provide electricity to rural users.  In
1996, federal utilities provided almost one-tenth of the nation's
power.  As a result of these activities, the federal agencies that
generate and/or market electricity and that make or guarantee loans
to finance improvements to rural electric systems had incurred a debt
of over $84 billion as of September 30, 1996.  This debt, it should
be noted, can be classified as direct and indirect.  The direct debt,
totaling over $53 billion, is owed directly to the federal
government--for example, RUS' borrowers owe about $32 billion.  The
indirect debt, over $31 billion, is owed by the federal agencies to
nonfederal parties--for example, TVA owed about $24 billion to
nonfederal bondholders. 


         FEDERAL AGENCIES GENERATE
         AND MARKET ELECTRICITY
------------------------------------------------------ Chapter 1:1.1.1

Federal entities that generate and/or market electricity--primarily
the Bureau, the Corps, the PMAs, and TVA--provided about 10 percent
of the nation's electricity supply in 1996.\1 The Bureau and the
Corps generate hydropower at about 130 federally owned power plants
located at federal water projects.  Because these projects are
managed for multiple purposes (for example, providing water for
irrigation, water supplies, navigation, flood control, and
recreation), the amount of power generated and marketed is affected
by the availability and use of water for these other purposes.\2

Power generated by the Bureau and the Corps is marketed by four of
DOE's five PMAs:  the Bonneville Power Administration (Bonneville),
plus the three that are the focus of this report:  the Southeastern
Power Administration (Southeastern), the Southwestern Power
Administration (Southwestern), and the Western Area Power
Administration (Western).  The fifth PMA, the Alaska Power
Administration, differs from the others in that it operates its own
power plants and distributes power directly to end-use (retail)
customers.\3 The PMAs in 1996\4 provided about 5 percent of the
nation's power. 

The PMAs' mission is to market federal hydropower at the lowest
possible rates that are consistent with sound business practices. 
The power the PMAs market is the power that remains after it has been
consumed for project purposes--for example, to pump water to fields
that are being irrigated.  By law, the PMAs are to give priority in
the sale of power to "preference customers"--public bodies (such as
municipal utilities, irrigation districts, military installations,
and other federal agencies) and cooperatives.  Each PMA has its own
specific geographic boundaries, federal water projects from which it
markets power, statutory responsibilities, and operation and
maintenance responsibilities.  Except for the Alaska Power
Administration, the PMAs generally do not own, operate, or control
the facilities that generate electric power; the generating
facilities are controlled by the operating agencies--most often the
Bureau and the Corps.  The PMAs, except for Southeastern, do own and
operate transmission facilities.  Southeastern relies on the
transmission services of other utilities to transmit the power it
sells to its customers. 

The PMAs are generally required to recover all costs incurred as a
result of producing, transmitting, and marketing power, including
repayment of the federal investment in the power generating
facilities and other debt, with interest.  Certain nonpower costs are
also allocated to power revenues for repayment.  For example, under
the concept of aid-to-irrigation, revenues earned from the sale of
power repay the federal investment in irrigation facilities that the
Secretary of the Interior deems is beyond the ability of irrigators
to repay.  According to Bureau officials, power revenues are
ultimately expected to cover about 70 percent of the federal
investment in completed irrigation facilities.  As of September 30,
1996, the PMAs and TVA had an outstanding debt of about $52 billion
related to financing the construction and operation of power plants,
transmission lines, and related electricity assets, as well as other
costs that are allocated to be repaid through revenues earned from
the sale of electricity.  TVA owed about $28 billion; Bonneville owed
about $17 billion; and Southeastern, Southwestern, and Western owed
the balance--about $7 billion.\5

Together, DOE's five PMAs and TVA\6 market power within 34 states. 
They do not serve Hawaii and states in the Northeast and upper
Midwest.  Figure 1.1 shows the service areas of the PMAs. 

   Figure 1.1:  The Service Areas
   of the PMAs

   (See figure in printed
   edition.)

Source:  Developed by GAO from data provided by DOE and the PMAs. 

The Congress established the first PMA, Bonneville, by passing the
Bonneville Project Act of 1937 to market federal power in the Pacific
Northwest.  (See app.  III for a more detailed discussion of
Bonneville.) In 1943, the Secretary of the Interior established
Southwestern under the President's war powers.  The Flood Control Act
of 1944 provided the authority to create PMAs and also gave the
Secretary of the Interior jurisdiction over the Corps' electric power
sales.  The Secretary of the Interior established Southeastern in
1950 and Alaska in 1967.  The last PMA, Western, was authorized by
the Department of Energy Organization Act of 1977, when the four
existing PMAs were transferred from the Department of the Interior to
DOE.\7

The largest individual federal power producer, however, is TVA, which
by some measures is the largest utility in the nation.  Providing
about 5 percent of the nation's power, TVA generates its own power
and markets it in wholesale markets, as well as directly to large
industrial customers.\8 TVA also approves the retail rates charged by
the 159 municipal and cooperative utilities that are its primary
customers.  In 1933, the Congress created TVA as a multipurpose,
independent federal corporation to develop the resources of the
economically depressed Tennessee River Valley:  TVA was to improve
navigation, promote regional agricultural and economic development,
and control the flood waters of the Tennessee River.  To those ends,
TVA erected dams and hydroelectric power facilities on the Tennessee
River and its tributaries.  Today, the power program is by far TVA's
largest activity, with about $5.7 billion in annual operating
revenues in fiscal year 1996.  TVA's hydroelectric facilities,
coal-fired power plants, nuclear generating plants, and other power
facilities--with a total generating capacity of over 28,000 megawatts
(MW)--provide electricity to nearly 8 million people in Tennessee and
parts of Alabama, Georgia, Kentucky, Mississippi, North Carolina, and
Virginia.  (See app.  I for a more detailed discussion of TVA.)


--------------------
\1 The latest year for which the PMAs provided this information at
the time we performed our review. 

\2 The evolution of the multiple purposes for federal water projects
is discussed in Bureau of Reclamation:  Reclamation Law and the
Allocation of Construction Costs for Federal Water Projects
(GAO/T-RCED-97-150, May 6, 1997). 

\3 The Alaska Power Administration's projects do not serve multiple
purposes the way other federal water projects do.  Its projects
provide power only.  Its power assets are being divested under the
Alaska Power Administration Sale and Termination Act, enacted in
November 1995.  DOE expects final divestiture by August 1998.  See
Federal Electric Power:  Views on the Sale of Alaska Power
Administration Hydropower Assets (GAO/RCED-90-93, Feb.  22, 1990). 

\4 The latest year for which the PMAs provided this information at
the time we performed our review. 

\5 See Federal Electricity Activities:  The Federal Government's Net
Cost and Potential for Future Losses:  Volume 1 (GAO/AIMD-97-110,
Sept.  19, 1997). 

\6 TVA markets power in Tennessee, as well as parts of Alabama,
Georgia, Kentucky, Mississippi, North Carolina, and Virginia. 
Southeastern, which sells power to TVA, also markets power within
these states, as well as other states in the Southeast. 

\7 The DOE Organization Act transferred power marketing
responsibilities and transmission assets that had been previously
managed by the Bureau of Reclamation to Western. 

\8 TVA sold power to 67 directly served industrial customers and
federal agencies in 1996. 


         THE GOVERNMENT ALSO MAKES
         OR GUARANTEES LOANS TO
         FINANCE THE CONSTRUCTION
         AND OPERATION OF RURAL
         ELECTRICITY SYSTEMS
------------------------------------------------------ Chapter 1:1.1.2

In addition to authorizing the sale of federal power in rural areas,
the Congress passed laws to encourage the development of nonfederal
power systems.  IOUs were historically reluctant to serve sparsely
populated areas because of the heavy capital costs involved in
installing power systems and serving relatively few customers.  As a
result, in 1935, scarcely 1 in 10 farm households in the United
States had electricity.  The Rural Electrification Act of 1936
authorized the Rural Electrification Administration (now RUS) to
provide loans and credit assistance to organizations that generate,
transmit, and/or distribute electricity to small rural communities
and farms.  From fiscal years 1992 through 1996, RUS made or
guaranteed 880 loans to rural utilities, some of which buy power from
the PMAs.  The outstanding balance on RUS' loans and loan guarantees
was about $32 billion as of September 30, 1996.\9 (See app.  II for a
more detailed discussion of RUS.)


--------------------
\9 The outstanding balance on RUS' loans and loan guarantees was
about $31 billion as of June 30, 1997. 


   NEW LEGISLATION AND
   TECHNOLOGIES SERVE AS A
   CATALYST FOR CHANGE IN
   ELECTRICITY MARKETS
---------------------------------------------------------- Chapter 1:2

From 1935 through the mid-1960s, little change occurred in the way
utilities satisfied demand for electricity and were regulated.  For
decades, they were able to meet increasing demand at decreasing
prices because they achieved economies of scale through capacity
additions and technological advances.  During much of this period,
demand for electricity grew at a faster rate than the gross national
product.  However, in 1976, electricity growth did not exceed overall
economic growth, and in 1982 electricity consumption declined.  These
adverse trends for the electric utility industry were caused by such
events as (1) the Northeast power blackout of 1965, which raised
concerns about reliability; (2) the Arab oil embargoes of the 1970s,
which resulted in increases in fossil fuel prices; and (3) the
passage of the Clean Air Act of 1970 and its 1977 amendments, which
required utilities to reduce pollutant emissions.  Because of the
decline in the rate of growth in demand for electricity, utilities
could no longer assume that prior patterns in demand-growth would
continue into the future.  How to satisfy the future demand for power
became an increasingly uncertain issue. 

In addition, since the late 1970s, statutory and technological
changes have created a climate for change in traditional electricity
markets.  In general, electricity markets are starting to evolve from
domination by large, monopolistic IOUs to competition among IOUs,
nonutility generators, power marketers, and others.  In the future,
electricity markets may evolve into ones in which electricity is a
commodity.  In addition, states are taking action to ensure that
retail consumers will be able buy power from a variety of competing
sources. 


      FEDERAL LAWS ENCOURAGE
      COMPETITION
-------------------------------------------------------- Chapter 1:2.1

In 1978, the Public Utility Regulatory Policies Act and the Fuel Use
Act encouraged the growth of a nonutility sector of the electricity
business.  These laws were passed to lessen the nation's dependence
on foreign oil and encourage alternative sources of power.  The
Public Utility Regulatory Policies Act required commercial utilities
to buy power from nonutility generators, called "qualifying
facilities." These entities had to meet certain criteria specified by
FERC for such matters as their ownership and operating efficiency. 
In addition, the act introduced the pricing of electricity on a
competitive basis:  As more nonutility generators entered the market,
FERC began approving certain wholesale transactions that had rates
that resulted from a competitive bidding process.  Many of the
qualifying facilities generated power in nontraditional ways--for
instance, by using small hydropower plants, cogeneration,\10 or
renewable sources.  Under the Fuel Use Act, electric utilities could
not use natural gas to fuel new generating technology; however, these
"qualifying facilities" could.  They were able to take advantage of
new generating technologies, such as combined-cycle gas turbine
generation\11 that can be built with less capital than larger power
plants.  Although the Fuel Use Act was repealed in 1987, qualifying
facilities and small power producers had already gained a portion of
the total electricity supply.  For instance, according to the
association of IOUs, in 1995 nonutility generators built about 60
percent of the nation's new electric generating capacity. 

The Energy Policy Act of 1992 was perhaps the most significant
legislative catalyst for increased competition.  It expanded
nonutility markets by creating a new category of power
producers--"exempt wholesale generators." Like qualifying facilities,
exempt wholesale generators do not sell their power in retail markets
and own only very limited transmission facilities.  Although FERC
does not regulate exempt wholesale generators under the Public
Utility Regulatory Policies Act, it regulates most of them as public
utilities under the Federal Power Act.  Under FERC's regulations,
exempt wholesale generators may charge market-based rates if they and
their affiliates lack market power.  Unlike the requirement under the
Public Utility Regulatory Policies Act that utilities purchase power
sold by qualifying facilities, there is no federal mandate that
utilities buy exempt wholesale generators' power.  The Energy Policy
Act also allows FERC, upon application, to order wholesale
wheeling\12 of electricity if such an order does not, among other
things, unreasonably impair reliability.  It is now possible for a
municipal utility that is served by an IOU to seek cheaper power from
a neighboring utility.  The Energy Policy Act also authorized FERC to
set transmission rates at levels that permit the utilities to recover
all of the costs incurred in providing transmission services,
including legitimate, verifiable, and economic costs. 

In April 1996, pursuant to its authorities under the Federal Power
Act, FERC issued a ruling on transmission access.  Order 888 requires
public utilities that own, control, or operate facilities that
transmit electricity in interstate commerce to offer both
point-to-point and network transmission services under terms and
conditions that are comparable to those that they provide for
themselves.\13 Public utilities must offer those services through
open-access, nondiscriminatory transmission tariffs\14 containing
minimum terms and conditions.\15 In addition, Order 888 allows
utilities the opportunity to seek recovery of certain stranded
costs\16 from those customers wishing to leave their current supply
arrangements.  However, according to the Deputy Director, FERC's
Office of Electric Power Regulation, the open-access provisions of
Order 888 do not apply to the PMAs, among other entities.  Therefore,
FERC cannot order the PMAs to provide open transmission services on a
general basis.  Operating under its authority under the Federal Power
Act, FERC can order the PMAs to provide transmission only on a
case-by-case basis.  However, to facilitate a unified national
approach to open-access transmission, DOE directed its PMAs that have
transmission facilities to publish generally applicable open-access
transmission tariffs, including ancillary services, in a manner
comparable to the service tariffs and other measures required of
transmission owners and operators that are regulated under FERC's
final rule.  In December 1997, Southwestern and Western filed
open-access transmission service tariffs with FERC, pursuant to Order
888.  The tariffs are to govern future access to available electric
transmission and, according to DOE, are consistent with the tariffs
of other wholesale transmission providers.  Bonneville had filed its
tariffs earlier. 


--------------------
\10 Cogenerators sequentially or simultaneously produce electric
energy and another form of energy (such as heat or steam) using the
same fuel source. 

\11 Combined cycle gas turbines use waste heat boilers to capture
exhaust energy from steam generation. 

\12 To "wheel" is to use the transmission facilities of one system to
transmit power and energy by agreement of, and for, another system
for a charge.  Wholesale wheeling usually refers to transmission
service to utilities that resell power to end users; retail wheeling
refers to transmission service to end users.  The act specifically
prohibited FERC, however, from ordering retail wheeling directly to
an ultimate consumer. 

\13 For purposes of Order 888, FERC has the authority to order open
transmission access on a generalized basis to "public
utilities"--IOUs and electric cooperatives with transmission assets
that do not have loans from RUS, among others.  FERC's order does not
apply to publicly owned utilities (e.g., municipal utilities and
public utility districts), TVA, or the PMAs. 

\14 A tariff sets forth rates, terms, and conditions of transmission
service. 

\15 A second FERC rule, Order 889, known as the Open Access Same-Time
Information System rule, requires public utilities to establish
electronic systems to share information about available transmission
capacity.  The order also requires utilities to separate their
wholesale power marketing and transmission operation functions but
does not require corporate unbundling or divestiture of assets. 

\16 Stranded costs are investments or assets owned by regulated
utilities that are not likely to be competitive in a restructured
marketplace.  More specifically, FERC defines wholesale stranded
costs as any legitimate, prudent, and verifiable cost incurred by a
utility to provide service to a wholesale requirements customer or a
newly created wholesale power sales customer that subsequently
becomes, in whole or in part, an unbundled wholesale transmission
service customer of such utility.  Order 888 allows utilities to seek
recovery of wholesale stranded costs associated with wholesale power
requirements contracts executed on or before July 11, 1994.




      UTILITIES RESPOND TO
      INCREASED COMPETITION
-------------------------------------------------------- Chapter 1:2.2

In response to the uncertainties about how the electricity market
will change and how fast, utilities have begun to implement new
strategies to compete.  Some are acquiring other utilities or merging
with them.  After years of virtually no mergers, many mergers have
been completed or proposed since the Energy Policy Act was enacted in
October 1992.  For example, for IOUs alone, from October 1992 to
January 1998, over 40 mergers had been proposed and 17 had been
completed, according to the Edison Electric Institute--the national
trade association for IOUs.  Utilities are also restructuring
themselves and decreasing their operating costs through
reorganizations and layoffs.  Some utilities are changing how they
plan to satisfy future demand for electricity and changing the types
of resources they acquire.  Because of uncertainty about market
conditions, instead of continuing to plan to meet long-term load
forecasts, utilities are focusing more on meeting more immediate
demand for power.  Thus, utilities are now tending to buy resources
that are flexible and allow them to adapt quickly to changing market
conditions, such as smaller natural gas-fired power plants and
purchased power.  Utilities are also retiring power plants if they
believe those plants may become uneconomic after the industry is
restructured. 

In responding to competitive challenges, utilities are trying to
compete for the business of other utilities' wholesale customers and
defending their business with existing customers.  For example, as
cited in our 1995 TVA report, Virginia Power cut one wholesale
customer's rates by 5 percent to fend off the marketing efforts of a
neighboring utility.\17

Federal power suppliers have also taken actions to become more
competitive.  For example, after the departure of half of its
industrial load, TVA froze its rates from 1986 through 1997, although
a rate increase was approved for 1998.  Moreover, Western recently
announced a decrease of over 20 percent, effective October 1, 1997,
in the composite rates of power it markets from hydropower plants in
the Central Valley Project in California.  In addition, according to
DOE's Power Marketing Liaison Office, Western began a process in
fiscal year 1995 to restructure itself.  The goals of this program
included reducing federal and contractor staff from fiscal year 1994
levels by 24 percent, saving $25 million in costs annually, and
reducing Western's organizational units.  For its part, Southwestern
has adopted a program to reduce overhead costs by reducing targeted
administrative positions, reducing the number of managers and
supervisors, and eliminating one field office. 


--------------------
\17 Tennessee Valley Authority:  Financial Problems Raise Questions
About Long-term Viability (GAO/AIMD/RCED-95-134, Aug.  17, 1995). 


   SEVERAL FACTORS WILL AFFECT HOW
   FAST COMPETITIVE MARKETS EMERGE
---------------------------------------------------------- Chapter 1:3

Electricity markets are not yet fully competitive but are moving in
that direction.  Although markets for wholesale transactions are
becoming competitive, retail markets are still uncompetitive. 
Supporters of restructuring argue that markets will not be truly
competitive until both wholesale and retail markets are transformed. 
In addition, other issues that need to be resolved include deciding
(1) how stranded costs are to be recovered, (2) how electricity is to
be transmitted in competitive markets, (3) how electricity is priced
in these markets, and (4) how consumers at the retail level are to be
offered a choice of power suppliers.  Once restructuring is complete,
retail electricity rates may fall between 6 percent to 19 percent by
the year 2015, depending on the intensity of competition, among other
factors, according to DOE's EIA. 


      RECOVERY OF STRANDED COSTS
-------------------------------------------------------- Chapter 1:3.1

Arguably the most significant issue that policymakers will face is
how to recover the stranded costs associated mainly with building
large baseload power plants and other assets under the old regulatory
regimen.  IOUs erected large amounts of nuclear generating capacity
and entered into long-term purchased power contracts to serve
existing and future loads.  Under the traditional covenants between
IOUs and their regulators, the capital and operating costs associated
with those assets were recovered through rates.  Now, with power
generation costs dropping and prospects that competition will affect
market prices, these high-cost plants are becoming uneconomical and
the costs associated with them may be "stranded." Estimates of the
investment in such assets nationwide range from $10 billion to $500
billion.\18

The issue of how to recover stranded costs--that is, who should
pay--is being debated.  In addressing the recovery of stranded costs
in the context of retail competition, some states have proposed
"sharing the pain":  Utilities could recover or offset the stranded
costs by taking mitigating actions (for example, by implementing
accelerated depreciation of generating assets, writing off the book
value of stranded assets, adjusting dividends to investors, or
decreasing operating expenses); ratepayers could pay through rate
increases that regulators hope will be temporary; or bonds could be
sold to the public to pay off the stranded costs and to avoid rate
increases.  However, some consumer groups believe that since
utilities incurred the costs, they should bear the burden of
repayment.  For example, an attempt to securitize the costs of a
nuclear power plant failed in Connecticut's legislature because
opponents, including consumer groups, believed the issuance of bonds
amounted to a "bailout" of the utility.  Staffs of state public
utility commissions have argued that because IOUs incurred stranded
costs under the old regulatory compact, IOUs should be allowed to
recover at least some of these costs before they must charge market
prices for power.  How stranded costs are divided between utilities
and their ratepayers, the period of time allowed for their recovery,
and how much the recovery of stranded costs affects rates will
determine when retail markets become competitive and to what degree. 


--------------------
\18 For example, see EIA's Electricity Prices in a Competitive
Environment:  Marginal Cost Pricing of Generation Services and
Financial Status of Electric Utilities--A Preliminary Analysis
Through 2015 (DOE/EIA-0614, Aug.  1997).  EIA estimates that if
regulatory means are not found to mitigate stranded costs, then the
reduction in market value for existing generating assets could range
from $72 billion to $169 billion (1995 dollars), under moderate
competition scenarios.  If competition is "intense," the reduction in
value may be as great as $408 billion.  These reductions may be about
$30 billion to $40 billion less over a 2-year period, during which
electricity markets phase in retail competition, because regulated
rates would continue to contain the stranded costs. 


      TRANSMITTING POWER IN
      COMPETITIVE MARKETS
-------------------------------------------------------- Chapter 1:3.2

To promote competition, new methods must be found to transmit power. 
Under current transmission arrangements, wholesale customers
frequently do not find it economical to buy power from a distant
utility because it must be transmitted over the power lines of
intervening utilities, each of which adds a transmission or wheeling
tariff to the price of the power.  For example, in 1995 during our
review of the financial viability of TVA,\19 we found that although
an IOU in the Southeast offered power that was competitively priced,
transmitting it to TVA's customers through one intervening utility
might increase the price by about 10 percent, rendering its delivered
price uncompetitive.  In addition, according to DOE officials, some
of the power transmitted is lost over distances. 

To facilitate competitive transmission of power, many state
regulators and FERC are advocating the establishment of "independent
system operators" (ISO).  Utilities in a given geographic area would
transfer the operation of their transmission assets to an independent
party that would transmit electricity reliably, safely, and
efficiently in a nondiscriminatory fashion.  For example, California
has established an Independent System Operation Restructuring Trust
to award funding to parties that will assist in establishing an ISO
to begin providing service in 1998.  The PMAs are also participating
in the formation of ISOs.  For example, Western is negotiating with
other utilities in the Southwest to establish the Desert Southwest
Transmission and Reliability Operator (an ISO) as well as to
participate in the California ISO.  Concerns exist that such
arrangements may be problematic from legal and constitutional
viewpoints.  According to Western officials, however, in Western's
agreements with other utilities pertaining to the ISO, Western is
taking care to ensure that its obligations under federal law and its
contractual agreements with preference customers are protected.  For
example, Western officials believe that, under language provided by
the PMA and accepted by FERC on Western's participation in the
California ISO, nothing in the ISO's tariff shall compel any person
or federal entity to violate federal statutes or regulations or
compel any federal agency to exceed its statutory authority as
defined in applicable federal statutes, regulations, or orders
lawfully promulgated thereunder.  These provisions also state that if
any provision of the tariff requires any person or federal entity to
give an indemnity or impose a sanction that is unenforceable against
a federal entity, the ISO shall submit to the Secretary of Energy or
DOE official a report of the situation.  The Secretary or other
official will take the steps necessary to remedy the situation to the
extent possible. 


--------------------
\19 GAO/AIMD/RCED-95-134. 


      PRICING POWER IN COMPETITIVE
      MARKETS
-------------------------------------------------------- Chapter 1:3.3

State public utility commissions are also taking steps to facilitate
competitive pricing of power.  They have supported establishing power
pools or exchanges.  Under these arrangements, members buy and sell
power through the pool or exchange it at a price that reflects market
demand and that promotes competition between utilities and other
suppliers.  For example, under one method, generating companies could
bid to sell their power to the pool.  The pool would then establish
hourly or spot prices based on these bids.  In California, the power
pool will publish prices every hour or half hour, to be viewed by
electric customers, investors, and power marketers.  With these
visible price signals, wholesale and retail buyers will be able to
make efficient purchasing decisions and adjust their consumption of
power from peak to off-peak periods when prices drop. 


      PROMOTING RETAIL COMPETITION
-------------------------------------------------------- Chapter 1:3.4

As of February 1998, all 50 states and the District of Columbia had
considered reforming their respective retail markets, according to
the National Regulatory Research Institute\20 and records obtained
from state regulatory agencies.  At that time, at least 17 states had
actually implemented plans to restructure the industry by enacting
restructuring legislation or by adopting final orders.\21 Regulators
in these states hope that industrial, commercial, and ultimately
residential consumers will be able to choose their power supplier,
rather than being tied to one utility.  These states hope to
establish retail choice at all levels by 1998 at the earliest and
2005 at the latest.  Supporters of retail competition hope that it
will nearly complete the restructuring process for electricity
markets and foster competitive pricing throughout the nation. 

At the time we completed our review, states such as Montana, New
Hampshire, and New York had asked utilities to implement pilot retail
choice programs so that broad issues that could affect widespread
competition later could be identified.  Several states, such as
Michigan, Pennsylvania, and Rhode Island, were implementing retail
competition in phases--for instance, extending it first to industrial
and commercial customers and then to residential customers.  As
mentioned previously, some states were addressing the issue of
stranded cost recovery.  In addition, at least 8 of these 17 states
were also encouraging utilities to continue their "social"
programs--such as energy efficiency and conservation programs, use of
renewable sources of power, and low-income energy assistance
programs.  These programs can be funded by charging consumers a
nonbypassable fee or by instituting a tax or surcharge on all energy
services.  Also, to foster competition and decrease utilities' market
power, public utility commissions were requiring utilities to
"unbundle" their services--that is, to divest themselves of, or
otherwise transfer, the generation, transmission, and distribution of
power. 

When restructuring is completed, states expect that retail customers
will enjoy a variety of options for taking advantage of retail
competition.  For instance, the California Public Utility Commission
expects that customers will use metered information about how much
power they are using at specific times of day and how much that power
costs.  They could then decide which supplier to buy from during
specific times to minimize costs.  They may be able to negotiate
directly with a supplier or use the services of an energy marketer or
broker.  In Maine, it is envisioned that consumers that are unwilling
to shop for alternative suppliers will be able to adopt the "standard
service option" from their existing utility.  The existing utility
will use a competitive bidding process in order to buy power for its
ratepayers at prices that are comparable to today's prices.  Other
options envisioned for Maine's ratepayers include signing contracts
with power marketers or aggregators that are short term, thus
enabling them to buy power at a low price but with a risk of rate
hikes or rate instability.  They will also be able to buy power under
longer term contracts at more expensive but more stable rates. 
Ratepayers will also be able to purchase "green power" (i.e., power
from nonpolluting sources such as renewable sources). 

Some states, however, are urging a cautious approach to retail
restructuring.  For example, the staff of Virginia's public utility
commission in an October 1996 report states: 

     "Those states that are aggressively pursuing competitive
     restructuring are invariably high-cost states with little to
     lose.  On the other hand, as a lower-cost state, Virginia may
     have little to gain and much to lose by being on the leading.  . 
     .edge of this restructuring movement.  We should also take note
     of the slow pace of those mostly low-cost states surrounding
     Virginia--North Carolina, Tennessee, Kentucky, West Virginia,
     and Maryland.  Consequently, Virginia should pursue a cautious
     and measured approach to adopting competitive initiatives, fully
     exploiting non-painful learning opportunities through observing
     the successes and failures of retail experiments and
     restructuring efforts in the more aggressive states."\22

Furthermore, in Nebraska, a state where all electric power is
provided by public entities and where power rates are among the
nation's lowest, the state's largest electric utility has asked a
federal appeals court to overturn FERC Orders 888 and 889.  The
utility challenged the orders on the grounds that FERC does not have
the legal authority to impose on the utility the same regulatory
regime that it imposes on private investor-owned electric utilities
because the utility is a political subdivision of the state of
Nebraska. 


--------------------
\20 The National Regulatory Research Institute was established by the
National Association of Regulatory Utility Commissioners to provide
research, educational, and technical services to the state regulatory
commissions. 

\21 The 10 states that had enacted legislation to restructure their
retail markets were California, Illinois, Maine, Massachusetts,
Montana, Nevada, New Hampshire, Oklahoma, Pennsylvania, and Rhode
Island.  Seven states that had adopted final orders without enacting
legislation were Arizona, Maryland, Michigan, New Jersey, New York,
Texas and Vermont. 

\22 Staff Investigation on the Restructuring of the Electric
Industry, Virginia State Corporation Commission (Oct.  1996). 


   OBJECTIVES, SCOPE, AND
   METHODOLOGY
---------------------------------------------------------- Chapter 1:4

Federal agencies that generate or market electricity and that make or
guarantee loans to finance improvements to rural power systems
incurred a debt of about $84 billion\23 as of September 30, 1996.\24
Three agencies that market federal electricity--the Southeastern,
Southwestern, and Western--are responsible for $7 billion of this
debt.  They face an uncertain future as electricity markets become
increasingly competitive.  In response, the Chairmen of the House
Committee on Resources and the Subcommittee on Water and Power asked
GAO to focus on these three PMAs and to (1) examine whether the
government operates them and the related electric power assets in a
businesslike manner that recovers the federal government's capital
investment in those assets and the costs of operating and maintaining
them and (2) identify options that the Congress and other
policymakers can pursue to address concerns about the role of these
three PMAs in restructuring markets or to manage them in a more
businesslike fashion.  GAO's options also apply to the Corps and the
Bureau, which generate most of the power these PMAs market.  Although
GAO's options apply only to these agencies, the report also provides
information about TVA, RUS, and Bonneville in appendixes I, II, and
III, respectively. 

We also included in this report information from generalized reports
on how federal agencies can be operated in a more businesslike
fashion.  See Related GAO Products at the end of this report for a
list of the products used to prepare this report. 

We conducted our review from April 1997 through February 1998 in
accordance with generally accepted government auditing standards. 
Appendix IV contains a detailed description of our objectives, scope,
and methodology. 

We provided a draft of this report to DOE's Power Marketing Liaison
Office that represented the views of Southeastern, Southwestern, and
Western; the Department of the Interior (including the Bureau); the
Department of Defense (including the Corps); Bonneville; and FERC. 
Their comments and our responses are included in appendixes VI, VII,
VIII, IX, and X, respectively. 


--------------------
\23 Dollar figures are in constant 1996 dollars, unless otherwise
specified. 

\24 GAO/AIMD-97-110. 


THE FEDERAL POWER PROGRAM OPERATES
AT A NET COST, HAS GENERATING
ASSETS THAT NEED REPAIR, AND
PRESENTS SOME RISK THAT THE
FEDERAL INVESTMENT MAY NOT BE
RECOVERED
============================================================ Chapter 2

Federal laws and regulations generally require that the PMAs recover
the full costs of producing and marketing federal hydropower.  The
PMAs generally follow these laws and regulations; however, in some
cases federal statutes and DOE's rules also prohibit or are ambiguous
about the recovery of certain costs.  As we reported in September
1997, for fiscal years 1992 through 1996, as a result of its
involvment in the electricity-related activities of Southeastern,
Southwestern, and Western (the three PMAs), the federal government
incurred "net costs" of $1.5 billion\1 --the amount by which the full
costs of providing electric power exceeded the revenues from the sale
of power.  In addition, the availability of many federal power plants
to generate electricity is below that of nonfederal plants because,
among other factors, the federal plants are aging and because the
federal planning and budgeting practices, including those used by the
Bureau and the Corps, do not always ensure that funds are available
so that repairs can be made when they are needed.\2 The resulting
declines in performance decrease the marketability of federal power. 
The net cost to the Treasury and the performance problems of the
federal power plants--when combined with competitive pressures on
electricity suppliers to decrease their rates at a time when some
federal hydropower project's environmental costs need to be recouped
by the PMAs--create varying degrees of risk that some of the federal
investment at certain federal generation and transmission projects
and rate-setting systems will not be repaid.\3 For example, although
the recovery of most of the federal investment in the three PMAs'
hydropower-related facilities is relatively secure, up to $1.4
billion of the federal investment for projects or rate-setting
systems pertaining to these PMAs, out of a total federal investment
of about $7.2 billion, is at some risk of nonrecovery. 


--------------------
\1 Dollars for the net costs are constant 1996 dollars, unless
otherwise specified.. 

\2 See, for example, Federal Power:  Outages Reduce the Reliability
of Hydroelectric Power Plants in the Southeast (GAO/T-RCED-96-180,
July 25, 1996). 

\3 A rate-setting system is a collection of one of more power
projects for which the PMAs set rates. 


   THE FEDERAL PROGRAM DOES NOT
   RECOVER ALL OF THE COSTS OF
   GENERATING, TRANSMITTING, AND
   MARKETING POWER
---------------------------------------------------------- Chapter 2:1

As noted in two of our recent products, the revenues of the
government's power generating and marketing activities are not
recovering all of the costs associated with the program.\4 These
activities operate at a net cost (loss) to the U.S.  Treasury.\5 For
the three PMAs that are the focus of this report, net costs of $1.5
billion were incurred for fiscal years 1992 through 1996.\6 These net
costs fall into several categories:  (1) net financing costs, (2)
unrecovered employee benefits, (3) unrecovered construction costs,
and (4) other costs. 


--------------------
\4 Power Marketing Administrations:  Cost Recovery, Financing, and
Comparison to Nonfederal Utilities (GAO/AIMD-96-145, Sept.  19, 1996)
and GAO/AIMD-97-110. 

\5 The government's power generating and marketing activities are
conducted by the Bureau, the Corps, Bonneville, the three PMAs, and
TVA.  The total net cost of these activities was about $3.6 billion
for fiscal years 1992 through 1996.  Bonneville's net costs were
about $2.1 billion--stemming from a net financing cost to the
Treasury of about $2 billion and unrecovered employee benefits of
about $110 million.  Moreover, TVA had net costs of about $4 million
because of unrecovered employee benefits.  (See apps.  I and III.)
Our totals exclude costs related to the Alaska Power Administration,
which is to be sold.  In addition to the $3.6 billion, the activities
of RUS in lending or guaranteeing loans to rural utilities incurred a
net cost of $4.9 billion during these years. 

\6 Dollars for the net costs are in constant 1996 dollars, unless
otherwise specified. 


      NET FINANCING COSTS OF THE
      THREE PMAS
-------------------------------------------------------- Chapter 2:1.1

We estimate that the net financing costs for the three PMAs'
appropriated debt\7 in fiscal years 1992 through 1996 was about $1.2
billion, including $208 million in fiscal year 1996.  These costs
stem primarily from appropriated debt provided by the federal
government at low interest rates with favorable repayment terms. 
Appropriated debt carries a fixed interest rate and cannot currently
be refinanced.  Also, the Treasury cannot require the PMAs to repay
the debt before it matures.  The interest the PMAs pay on their
outstanding appropriated debt is often substantially below the rate
the Treasury incurred to provide funding to the PMAs.  The PMAs'
average interest rate on outstanding debt was 3.5 percent,\8 whereas
the Treasury's weighted average interest rate on outstanding bonds
was 9 percent\9 to provide funding to the PMAs.  The PMAs have
incurred substantial amounts of appropriated debt at low interest
rates primarily because, in accordance with the appropriate DOE
order,\10 they repay high-interest debt first, and because the
appropriated debt they incurred before 1983 was generally at the
below-market interest rates in effect at the time.\11


--------------------
\7 We use the term "appropriated debt" because the PMAs are required
to repay appropriations used for capital investments, with interest. 
However, the Treasury does not technically consider these
reimbursable appropriations to be lending. 

\8 Because audited fiscal year 1996 data were not available for
Southeastern and Southwestern at the time of our fieldwork, we used
fiscal year 1995 appropriated debt and weighted average interest
rates.  According to the PMAs, the appropriated debt balances did not
change significantly in fiscal year 1996.  We then calculated fiscal
year 1996 net financing costs using the 1996 Treasury average
interest rate. 

\9 This rate is the weighted average interest rate on the Treasury's
entire outstanding bond portfolio (10- to 30-year maturities).  We
used this interest rate because it reflects Treasury's average
interest rate on outstanding long-term debt and most closely matches
the terms of the PMAs' appropriated debt. 

\10 DOE Order RA 6120.2, "Power Marketing Administration Financial
Reporting," generally requires the PMAs to repay the highest interest
rate debt first, while complying with the repayment periods and
unless otherwise indicated by legislation. 

\11 In 1983, DOE required that, absent specific legislation to the
contrary, appropriations for capital expenditures made after
September 30, 1983, would be financed at interest rates equal to the
average yield during the preceding fiscal year on interest-bearing
securities of the United States, which, at the time the computation
was made, have terms of 15 years or more remaining before maturity. 


      PMAS' RATES DO NOT RECOVER
      ALL EMPLOYEE BENEFIT COSTS
-------------------------------------------------------- Chapter 2:1.2

For current PMA and operating agency employees, the federal
government incurs a portion of the cost for Civil Service Retirement
System pensions and almost all of the cost for postretirement health
benefits.  For fiscal years 1992 through 1996, we estimate that the
net cost to the federal government of providing these benefits was
about $82 million for the three PMAs, including $16 million in fiscal
year 1996.  The PMAs plan to begin recovering the full annual cost of
pension and postretirement health benefits in fiscal year 1998.\12


--------------------
\12 Consistent with current policies and laws, the PMAs do not plan
to recover pre-1998 costs. 


      PMAS' RATES DO NOT RECOVER
      ALL CONSTRUCTION COSTS
-------------------------------------------------------- Chapter 2:1.3

We found that the three PMAs had incurred costs or had costs
allocated to them by the operating agencies for which full costs were
not being recovered through the PMAs' rates.  These costs were for
the few projects that were not yet completed, were under
construction, or were canceled.  In some cases, this situation
occurred because the power generating projects had never operated as
designed.  In accordance with DOE's guidance, the PMAs set rates that
exclude the costs of nonoperational parts of the power projects,
including capitalized interest.  For example, at the Corps' Russell
Project (located on the Savannah River, which serves as the border
between Georgia and South Carolina), partially on line since 1985,
litigation over large fish kills has kept four of the eight turbines
from becoming operational.  As a result, over half of the project's
construction costs--about $500 million--have been excluded from
Southeastern's rates.  The net costs of these construction projects
for fiscal year 1996 represent capitalized or unpaid interest
incurred in that year.  For construction projects designed to
generate power marketed by the three PMAs, we estimate that for
fiscal years 1992 through 1996, the cumulative net costs are $138
million, including $30 million in 1996.  The PMAs believe that in
most instances, including the Russell project, these net costs will
be recovered in future years. 


      PMAS' RATES DO NOT RECOVER
      SOME OTHER COSTS
-------------------------------------------------------- Chapter 2:1.4

The three PMAs also incurred other net costs that totaled $157
million for fiscal years 1992 through 1996, for such purposes as
environmental mitigation and irrigation.  In an example involving
environmental mitigation, at the Central Valley Project's Shasta Dam
in California, the 1991 Energy and Water Development Appropriations
Act specified that any increases in Western's costs to purchase power
because of bypass releases to preserve fisheries downstream should
not be allocated to power; instead, they were paid for by
appropriated funding.  These costs totaled about $15.3 million in
fiscal year 1996 and about $53.8 million for fiscal years 1992
through 1996.\13 In another example of net costs related to
irrigation, in May 1996 we estimated that about $454 million in (1)
the federal investment in hydropower facilities allocated to
irrigation at the Bureau's Pick-Sloan Missouri Basin Program\14 and
(2) a portion of the costs associated with storing water for these
projects were not likely to be recovered without congressional
action.\15

The principal of $454 million had grown to $464 million as of
September 30, 1996.  As, by law, interest on this amount is not being
paid, we estimated that about $70.6 million in interest was unpaid
for fiscal years 1992 through 1996. 


--------------------
\13 According to DOE's Power Marketing Liaison Office, the costs
incurred by Western for Shasta bypasses totaled only $1.9 million in
fiscal year 1997.  Also, as of September 30, 1997, all future
purchased power costs incurred by Western due to cold water releases
at the Shasta Dam will be reimbursable or included in the power rates
for repayment purposes. 

\14 The Program consists of 13 of the Corps' and the Bureau's
hydropower plants and associated irrigation projects, among other
assets, located in the northern basin of the Missouri River.  Western
sets rates that are designed to recover, not only the federal capital
investment in the power system, but also part of the federal
investment in irrigation, as well as other costs. 

\15 Federal Power:  Recovery of Federal Investment in Hydropower
Facilities in the Pick-Sloan Program (GAO/T-RCED-96-142, May 2,
1996). 


   THE FEDERAL HYDROPOWER ASSETS
   NEED REPAIR
---------------------------------------------------------- Chapter 2:2

The availability of federal power plants to generate power is below
that of other power plants.  Many federal plants are aging (the
Bureau's plants average about 50 years in service and the Corps'
about 30 years), which increases the need for repairs.  At the same
time, the Bureau's and the Corps' planning and budgeting processes do
not always provide funding to repair the federal power assets when
the funding is needed, causing some repairs to be delayed and the
power plants to become less available to provide power. 

According to the representatives of the PMAs' power customers and our
previous work, the maintenance needs of the Bureau's and the Corps'
hydropower plants are often underfunded or maintenance is delayed. 
Furthermore, data from both operating agencies show that their power
plants are generally less available to generate power than power
plants operated by other generators of electricity.  For example,
according to the Bureau's 1996 benchmarking study, while the agency's
power plants exceeded the performance of the industry in terms of
wholesale firm rate, production costs/kWh, and the number of
full-time operation and maintenance employees per generating unit,
they lagged behind other nonfederal and federal hydropower producers
in availability, forced outage, and scheduled outage factors.\16
However, the availability of the Bureau's hydropower plants over the
last 3 years has been above the average availability of the last 15
years.  In our 1996 testimony,\17 we reported that in the Corps'
South Atlantic Division, the availability of hydropower plants
declined from about 95 percent in 1987 to 87 percent in 1995.  In
addition, the 1995 availability of the Corps' units is below the
industry average (89 percent availability) in the Bureau's
benchmarking study.  Several hydropower plants have been off line for
several years because of forced outages.\18 However, DOE's Power
Marketing Liaison Office notes that maintenance problems differ by
region, district, or division within the operating agencies and that
problems in one area should not be extrapolated to all areas. 

The planning and budgeting processes that federal agencies--including
the Bureau and the Corps--use are not conducive to predictable
planning and funding of needed repairs.  Pursuant to key laws,
including the Antideficiency Act, the Adequacy of Appropriations Act,
and the Budget Enforcement Act, federal agencies cannot enter into
obligations prior to an appropriation and cannot exceed
appropriations unless they have specific statutory authority to do
so.  Thus, they cannot enter into contracts that obligate them to pay
for goods or services unless sufficient funds are available to cover
the costs in full.  Therefore, agencies must budget for the full
costs of contracts up front.  Agencies cannot enter into a contract
unless it is authorized by law and an appropriation covers the
contract's cost.  Moreover, fixed spending limits, or caps, apply to
all discretionary spending through 1998, including spending for
capital items.  As we reported in 1996, agency officials often
pointed to the poor condition of federal power plants as evidence of
a need for more capital spending and reformed budgeting.\19 Some
observers add that increased capital spending is needed to generate
operational savings in the future.  They believe that in an era of
constrained federal budgets, spending on capital projects is limited
because it entails heavy initial costs and the budget "scoring" for
such projects occurs in a single year, while the benefits of it
extend for many years. 

PMAs and their customers stated that they view the federal planning
and budgeting processes as not being well adapted to a commercial
activity, such as operating a power system.  Under current planning
and budgeting systems, the project and field locations of the Bureau
and the Corps identify, estimate the costs of, and develop their
budget proposals, not only for hydropower but also for such
facilities as dams, irrigation systems, and recreational facilities. 
Hydropower repairs may be assigned lower priorities than other items. 
Budget requests also have been subject to 10-percent to 15-percent
reduction targets at the operating agencies.  Under these conditions,
the operating agencies, the PMAs, and the PMAs' hydropower customers
believe that funding for needed repairs is at best uncertain and at
times is unavailable when needed.  To ensure that the funding of
hydropower maintenance and repair activity receives the funding
priority they believe it deserves, customer groups are encouraging
the operating agencies to consult them about budgeting and planning
for operation and maintenance.  Customer groups are also encouraging
the federal agencies to seek alternative funding.  In most cases, the
customers are willing to provide up-front financing for repairs if
they are granted more input to planning and budgeting decisions,
according to DOE's Power Marketing Liaison Office.\20


--------------------
\16 Bureau of Reclamation, "Future Generations:  A New Era of Power,
Performance, and Progress," 1996.  According to this document, the
Bureau's plants were available about 83 percent of the time in 1994,
compared with an industry average of 89 percent. 

\17 GAO/T-RCED-96-180. 

\18 According to data provided the Corps, agencywide its hydropower
plants were available to provide power 92.9 percent of the time in
fiscal year 1987 but only 87.9 percent of the time in fiscal year
1995.  However, the availability of these plants improved to 88.4
percent in fiscal year 1996 and 89 percent in fiscal year 1997. 
Also, the percentage of the Corps' plants that experienced forced
outages decreased from 5.98 percent in fiscal year 1995 to 4.44
percent in fiscal year 1997.  The Corps attributes these
improvements, in part, to its program to spend $450 million on
repairing its hydropower assets from fiscal years 1993 to 2007. 

\19 Budget Issues:  Budgeting for Federal Capital (GAO/AIMD-97-5,
Nov.  12, 1996). 

\20 These customers acknowledge that although they can advise the
Bureau, the Corps, and the PMAs about capital improvements to be
undertaken and the levels of funding needed, the federal agencies
retain the ultimate decision-making authority and continue to own the
facilities. 


   RISK OF NONRECOVERY OF SOME
   FEDERAL INVESTMENT EXISTS
---------------------------------------------------------- Chapter 2:3

In our September 1997 report, we found that the risk exists that some
portion of the government's investment in its power generation and
sales program may not be recovered.\21 The total amount of investment
in the assets of the power generating and marketing programs of the
operating agencies, the three PMAs, Bonneville, and TVA was about $52
billion.  This risk stems from several factors, two of which have
been addressed already in this report.  First, the large net costs of
the federal hydropower program will continue if action is not taken
to recover all of the costs of operating the program.  Second, the
degraded availability of the generating assets contributes to this
risk of nonrecovery because it decreases the marketability of federal
power.  Other factors also add to the risk of nonrecovery.  One
factor is that the onset of market competition puts pressure on
suppliers to keep their electric rates low or to decrease them.  At
the same time, the PMAs are being pressured to raise some rates
because of the costs at certain projects for mitigating the damage to
fish and wildlife habitat from hydropower generation.  Moreover, when
the operating agencies have had to curtail power generation at
particular projects to protect the environment, the PMAs have had to
purchase power to fulfill their contracts--another factor that puts
upward pressure on the PMAs' rates. 


--------------------
\21 GAO/AIMD-97-110. 


      TREND TOWARD LOWER MARKET
      RATES CREATES SOME RISK OF
      NONRECOVERY OF THE FEDERAL
      INVESTMENT
-------------------------------------------------------- Chapter 2:3.1

Nationwide electricity rates have dropped over 25 percent after
inflation since 1982.  According to DOE's Energy Information
Administration, retail rates fell from a nationwide average of 8.7
cents per kWh in 1982 to 6.3 cents in 1996 (constant 1992 dollars). 
This decrease has been caused by factors that include declining fuel
prices, an increasing number of fully depreciated power plants, more
efficient power generation, and competition from nonutility
generators.  According to various industry analysts, the
restructuring of electricity markets will cause market rates to
continue to decline.  In addition, according to the Energy
Information Administration, retail rates nationwide in 2014 may be
about 6 percent to 19 percent below the levels they would have been
if competition had not begun.  In some cases, wholesale power is
available today at about 2 cents per kWh.  For example, according to
the customer group of the Colorado River Storage Project, in May 1997
one Western customer signed a 20-year contract with an IOU to
purchase firm power at a rate not to exceed 1.8 cents per kWh.  In
contrast, Western's composite rate for power from the project was
about 2 cents per kWh.  If the PMAs' customers can buy less expensive
power from sources other than the PMA, the fixed costs associated
with the federal government's power assets will need to be recovered
from a decreasing number of customers, placing increased pressure on
the PMA to increase its rates.  This pressure, in turn, will
encourage additional customers to seek power from other sources. 


      ENVIRONMENTAL MITIGATION
      COSTS ALSO ADD TO THE RISK
      OF NONRECOVERY
-------------------------------------------------------- Chapter 2:3.2

At the same time that wholesale and retail rates are declining, the
PMAs are being pressed to raise rates at some projects, primarily
because of the need to address concerns about damages to the
environment and endangered species.  As a result, the three PMAs'
hydropower programs have lost revenues, have had to buy more costly
replacement power to fulfill their contacts with their power
customers, and in some cases have had to spend millions of dollars to
mitigate environmental effects.  For example, according to DOE's
Power Marketing Liaison Office, about one-third of the 1,356 MW
capacity at the Bureau's Glen Canyon Dam in Arizona, whose power is
marketed by Western, could be lost because power generation has been
restricted to protect recreational resources and endangered fish
species.  The Bureau estimates that Western has lost more than $100
million in revenues.  At the same time, Western's costs to buy power
to replace the lost generating capacity have averaged about $44
million per year.\22 Furthermore, at the Bureau's Shasta power plant,
in California, whose power Western also markets, restrictions on the
turbine operations and cold water bypasses to protect the winter run
of the chinook salmon resulted in about $50 million in additional
costs to purchase power for Western since 1987.\23 Moreover, the
shutdown of some units at the Corps' Russell project because of
litigation over fish kills resulted in Southeastern's losing $36.1
million in revenues per year since fiscal year 1994. 


--------------------
\22 Federal Power:  Issues Related to the Divestiture of Federal
Hydropower Resources (GAO/RCED-97-48, Mar.  31, 1997). 

\23 According to Bureau officials, the bypasses ceased by November
1996 because of the installation of temperature control devices. 
These devices cost $80 million, according to Western. 


      UP TO $1.4 BILLION OF
      FEDERAL INVESTMENT IN THE
      POWER ASSETS OF THE THREE
      PMAS IS AT RISK OF
      NONRECOVERY
-------------------------------------------------------- Chapter 2:3.3

As we recently reported, some portion of up to about $1.4 billion in
federal investment is at varying degrees of risk of not being
recovered through power revenues at three generation projects, one
transmission project, and two rate-setting systems pertaining to the
three PMAs.  As of September 30, 1996, the three PMAs had accumulated
over $7.2 billion in debt for constructing and upgrading the Bureau's
and Corps' generating facilities whose power the three PMAs market,
the PMAs' transmission facilities, and the Bureau's irrigation
facilities, which are largely repaid with power revenues.\24 In
general, the recovery of most of this investment is seen as
relatively secure because the three PMAs are generally competitively
sound:  Their cost to generate power, measured in terms of average
revenue per kWh, was 40 percent or more below nonfederal utilities
for 1995.  However, at some projects, congressional action will be
needed to ensure that large amounts of federal investment are
recovered.  For example, at the Pick-Sloan Program, $464 million in
federal investment in power facilities and reservoir storage cannot
be recovered until the associated irrigation projects come into
commercial service.  Because most of these irrigation projects are
infeasible, the $464 million cannot be repaid.  Without congressional
action to force a reallocation of these costs from irrigation to
power, or a related solution, recovery cannot take place.  Recovery
of these costs would place upward pressure on Western's electricity
rates--potentially entailing a one-time increase of up to 14.6
percent.  At a time that wholesale electric rates are decreasing,
such increases in the PMAs' rates are uncompetitive and could erode
the marketability of the federal power if they are numerous and
continuous.  Table 2.1 contains information about the circumstances
surrounding the $1.4 billion at risk.  Additional details on the
situations at these six projects or systems are presented in appendix
V. 




                                    Table 2.1
                     
                          Risk of Nonrecovery of Federal
                       Investment in Assets Associated with
                     Southeastern, Southwestern, and Western,
                             as of September 30, 1996

                              (Dollars in millions)

                                                      Dollar
                                                           s
                  Project/          Risk                  at
PMA               system            category\a          risk  Explanation
----------------  ----------------  ----------------  ------  ------------------
Southeastern      Russell project   Remote,             $518  Pumping units\b
                                    reasonably                (about 300 MW) are
                                    possible,                 inoperable because
                                    probable                  of litigation over
                                                              fish kills. As a
                                                              result, federal
                                                              capital investment
                                                              has not been
                                                              recovered through
                                                              rates. Risk of
                                                              loss is remote if
                                                              units are placed
                                                              into service,
                                                              reasonably
                                                              possible if
                                                              inclusion of costs
                                                              in rates makes the
                                                              rates
                                                              noncompetitive,
                                                              and probable if
                                                              these units do not
                                                              come on line.\c

Southwestern      Truman project    Remote, probable      31  Pumping capacity
                                                              is not functioning
                                                              because of design
                                                              flaws and
                                                              excessive fish
                                                              kills. Risk of
                                                              loss to government
                                                              is remote only if
                                                              the units are
                                                              placed into
                                                              service as
                                                              designed in the
                                                              near future. This
                                                              is an unlikely
                                                              event because they
                                                              have been off-
                                                              line since the
                                                              early 1980s.
                                                              Otherwise, the
                                                              risk is
                                                              probable.\d

Western           Central Valley    Reasonably           267  Some portion of
                  Project (rate-    possible                  the investment is
                  setting system)                             at risk for
                                                              nonrecovery,
                                                              mainly because
                                                              environmental
                                                              legislation
                                                              requires a
                                                              reallocation of
                                                              water among its
                                                              uses, which could
                                                              result in
                                                              restrictions on
                                                              its use to
                                                              generate power. At
                                                              the same time, the
                                                              Central Valley
                                                              Project's power is
                                                              faced with
                                                              competition from
                                                              nonfederal
                                                              generators.\e

Western           Pick-Sloan        Probable             464  The federal
                  Missouri Basin                              investment in
                  Program (rate-                              hydropower
                  setting system)                             capacity and
                                                              reservoir storage
                                                              originally
                                                              intended for use
                                                              by future
                                                              irrigation
                                                              projects will not
                                                              be repaid without
                                                              congressional
                                                              action. Under
                                                              program statutes,
                                                              recovery through
                                                              rates cannot occur
                                                              until the
                                                              irrigation
                                                              projects come into
                                                              commercial
                                                              service. According
                                                              to the Bureau,
                                                              almost all of
                                                              these projects are
                                                              infeasible.

Western           Washoe Project    Reasonably            13  Since January
                                    possible,                 1996, Western has
                                    probable                  estimated that to
                                                              cover Washoe's
                                                              annual operating
                                                              expenses
                                                              (excluding
                                                              depreciation),
                                                              interest charges,
                                                              and debt
                                                              repayment, power
                                                              from the project
                                                              would have to be
                                                              priced from 5.7
                                                              cents to 11 cents
                                                              per kWh. However,
                                                              Washoe's average
                                                              revenue per kWh
                                                              for energy sales
                                                              in 1996 was only
                                                              1.02 cents per
                                                              kWh. If Washoe's
                                                              power continues to
                                                              be marketed stand-
                                                              alone, losses are
                                                              probable, but they
                                                              are reasonably
                                                              possible if
                                                              Western blends
                                                              Washoe's rates
                                                              with those of the
                                                              Central Valley
                                                              Project.\f

Western           Mead-             Reasonably            95  Only about $71,000
                  Phoenix           possible,                 of a $95 million
                  (transmission     probable                  federal investment
                  project)                                    has been recovered
                                                              because demand had
                                                              not materialized
                                                              for power or
                                                              transmission
                                                              services. Losses
                                                              to the government
                                                              are probable if
                                                              the services are
                                                              marketed stand-
                                                              alone, but
                                                              reasonably
                                                              possible if they
                                                              are blended with
                                                              other systems.

Total                                                 $1,388
--------------------------------------------------------------------------------
\a Based on Statement of Federal Financial Accounting Standard No. 
5, Accounting for Liabilities of the Federal Government, if the
chance that a contingent loss will occur is more likely than not, the
loss is "probable"; if the chance is more than remote but less than
probable, it is "reasonably possible"; if the chance is slight, it is
"remote."

\b Pumping units are designed to allow water, after it has passed
through the generating units, to be pumped back into the reservoir
during periods when demand for power is low.  Then, the water can be
used to produce power during periods of higher demand. 

\c According to DOE's Power Marketing Liaison Office, some
unspecified portion of the $518 million investment in pumping units
at this project will be recovered even if the units are never
commercially operated.  However, we believe this assertion overlooks
the policy guidance contained in DOE Order RA 6120.2, which indicates
that if the nonoperational units are not placed into commercial
service, the power customers will not be required to repay the
investment. 

\d According to the Corps, all repairs to off-line generating units
will be completed by February 1999.  According to DOE's Power
Marketing Liaison Office, Southwestern can add to its power repayment
study the power-related costs of this project's pumpback units even
if the units are never operable.  The Corps' ability to use the units
in a pumping capability awaits the lifting of an injunction by the
state of Missouri.  However, we believe this assertion overlooks the
policy guidance contained in DOE Order RA 6120.2, which indicates
that if the nonoperational units are not placed into commercial
service, the power customers will not be required to repay the
investment. 

\e Western announced a decrease of over 20 percent, effective October
1, 1997, in the composite rates of power it markets from the Central
Valley Project.  FERC approved these rates on a final basis on
January 8, 1998.  These rate cuts were facilitated by renegotiating
contracts that obligate Western to purchase power for its customers
if the Project cannot supply enough power.  The sustainability of
these rate cuts, however, is uncertain, because of the effects of the
Central Valley Project Improvement Act.  Specifically, under the
act's provisions, 800,000 acre feet of water in the Project must be
managed for environmental purposes.  According to the Bureau, an
analysis of environmental impacts indicates that this change in how
water is managed may result in a 5 percent reduction in hydropower
production. 

\f According to DOE's Power Marketing Liaison Office, Western staff
are proposing the blending of the costs of power from this project
with the costs of the Central Valley Project after the year 2004. 

Source:  GAO/AIMD-97-110 and data provided by the Bureau, the Corps,
DOE's Power Marketing Liaison Office, and Western. 


--------------------
\24 Under the concept of aid-to-irrigation, power revenues are to pay
for the federal investment in irrigation facilities that the
Secretary of the Interior deems to be beyond the irrigators' ability
to repay. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 2:4

More competitive electricity markets will offer new benefits to
consumers while posing a special challenge to the federal
government's program to generate and market power.  With competition
at the wholesale and retail levels, ratepayers are likely to enjoy
unprecedented opportunities to choose from among several competing
suppliers offering a variety of prices and services.  However, the
problems we have reported in recent years, combined with these market
changes, should alert policymakers to take steps to protect the
investment in the federal power assets. 

Even in the absence of market changes, the agencies that provide
power are over $50 billion in debt, including about $7 billion for
the three PMAs.  At the same time, the hydropower assets are
degrading in terms of their availability to generate power, thereby
making the power they generate less marketable.  As competitive
markets develop, some PMA customers may opt to buy from other
suppliers if the PMAs' power is perceived as being increasingly
unreliable.  In addition, although the PMAs' power is very
competitively priced, this advantage may not last.  Specifically,
competition is expected to cause market rates to fall.  At the same
time, the PMAs' rates need to cover the costs of environmental
impacts downstream.  If the PMAs' rates increase and the wholesale
rates for power fall to the point where the two rates converge, the
PMAs may lose customers to other suppliers.  At the Central Valley
Project and the Colorado River Storage Project, Western's wholesale
power is already priced at levels competitors can challenge. 

If the PMAs lose customers to other suppliers, then the risk
increases that the federal investment in the power program will not
be recovered.  As documented in this chapter, for the three PMAs'
projects and rate-setting systems, some portion of $1.4 billion is
already at risk for nonrecovery.  Although most of the risk to the
$1.4 billion does not stem from increasing competition, the advent of
competition does heighten the risk of nonrecovery.  As discussed in
the next chapter, options are available to the Congress and the
agencies themselves to better recover costs and protect the federal
investment, among other benefits. 


OPTIONS FOR OPERATING FEDERAL
HYDROPOWER ASSETS
============================================================ Chapter 3

The nation's electricity markets are undergoing significant changes,
as the previous chapters have shown.  The speed with which this
widespread restructuring may be completed is uncertain; however, it
is ongoing and will continue, perhaps at an accelerating pace, as
proposals to expand competition to the retail electricity market
continue to be made by national and state policymakers, electric
utility interest groups, and the Congress.  As the industry becomes
less regulated and more competitive at both the wholesale and retail
levels, nonfederal utilities and power suppliers have taken important
steps to become competitive to survive.  Federal power agencies also
face the challenge of moving to a more competitive environment.  The
entities to whom the PMAs sell power, aware that they need to supply
the cheapest available power to their own retail customers, have
begun to pressure the PMAs, the Bureau, and the Corps to adopt
business practices that are better suited to the new era. 
Furthermore, and perhaps most important, these agencies are under
pressure to adapt to the new markets to reduce the risk that the
multibillion-dollar federal investment in hydropower and other
associated programs will not be repaid if federal power ultimately
proves to be too unreliable and overpriced to be competitive.  In
this connection, a widening recognition exists today that options for
operating federal hydropower assets need to be considered and
ultimately implemented.  Three broad options exist for addressing the
federal hydropower program's operations: 

  -- Preserve the status quo of federal ownership. 

  -- Maintain federal control of the hydropower assets but manage
     them in a more businesslike manner. 

  -- Divest the federal hydropower assets. 


   PRESERVE THE STATUS QUO
---------------------------------------------------------- Chapter 3:1

The federal power program uses low-cost hydropower generated at major
federal water projects to help meet the needs of the preference
customers, many of which are located in rural areas.  The power
plants at these water projects are generally operated by the Bureau
and the Corps--the operating agencies--and the power that exceeded
the project's operational requirements is marketed by the PMAs, as
described in chapter 1.  Power is generated and marketed in a way
that balances how the water is being used for the other purposes of
the projects.  Funding for the activities of the operating agencies
and the three PMAs is subject to the annual congressional
appropriation process under which the agencies obtain their funding
for capital investments as well as for operations and maintenance
expenses. 

PMA and operating agency officials and representatives of the PMAs'
customer associations have indicated a need to change how the federal
hydropower program is being operated.  They stated that the agencies'
planning and budgeting processes do not provide sufficient,
predictable, and timely funding to facilitate the repair of the
federal power plants.  In addition, they pointed to various
administrative and legal requirements that they believe cause the
PMAs and operating agencies to generate and market power in an
unbusinesslike manner.  In this connection, they have advocated ways
to manage the federal hydropower assets, discussed in the next
section, that will address these concerns. 

Some representatives of the PMAs' preference customers have advocated
defederalizing the PMAs and the federal generating assets as a way of
improving their operating efficiency and availability.  For example,
according to an official of an association of Western's municipal
power customers, the preference customers should purchase the federal
generating and transmission assets of the Colorado River Storage
Project in order to avoid the sharp rate increases that characterized
Western's rates from the project since the late 1980s.  It is
important to note, however, that other preference customers continue
to support continued federal ownership of the dams, reservoirs, and
hydropower assets.  These customers believe that, although some
changes in the PMAs' current practices could lower operating costs
and improve efficiencies, as a whole the PMAs have offered
high-quality, low-cost services while balancing the diverse needs of
the beneficiaries of the federal multi-use projects. 

Moreover, representatives of investor-owned utilities or proponents
of divestiture have questioned why the federal government continues
to provide power in restructuring markets.  First, electrifying rural
areas was an important goal of the federal power program; however,
this goal has been largely satisfied.  Therefore, the need for the
federal government's involvement is questionable.  Second,
competition likely would enable wholesale and retail customers to
choose from among competing power suppliers.  This possibility again
questions the need for the federal government to sell power.  Third,
the issue of providing low-cost PMA power to portions of 34 states in
the South and West where the preference customers of the PMAs are
located, but not to other areas, is debatable.  And fourth, IOUs and
other critics of PMA power state that, as federal agencies, the PMAs
have advantages that IOUs do not have and therefore would compete
with their nonfederal parties on an uneven basis.  For example, our
work has shown that the PMAs have rates that do not recover all of
the costs of generating, transmitting, and marketing power.  Also, as
federal agencies, the PMAs are not subject to income taxes or state
regulatory oversight and have more flexible repayment and
rate-setting methodologies.  Fifth, the status quo continues the
existing risk of nonrepayment of the federal investment. 

Because of the stakes involved in changing the management and
ownership of federal water projects and hydropower plants,
maintaining the status quo affords policymakers the opportunity to
make careful decisions about how to proceed.  The federal
government's role in balancing the multiple uses of water is
important.  It affects such things as how much water will be
available to accommodate the expansion of metropolitan areas, how
much water will be used to protect endangered species, and how much
water will be needed to protect the harvesting of shellfish in the
Apalachicola Bay, Florida.  The Bureau and the Corps generate power
while balancing these impacts.  Any decisions that federal
policymakers reach about changing how power is generated or how the
water projects will be managed or owned will need to consider the
impacts of the decisions on the uses of the water and the
beneficiaries of the water projects.  An advantage of the status quo
is that it continues the federal role in balancing the multiple uses
of the water and allows policymakers time to study these issues
before they change the operations and/or ownership of the water and
the power assets.  Also, by preserving the existing multiple uses of
the water projects and the projects' beneficiaries, the status quo
avoids the debate that is likely to occur if the Congress reexamines
the agreements reached decades ago on federal involvement in power. 
For example, the status quo continues federal power's role in helping
promote the economies of rural areas, especially by providing
inexpensive power to these areas for homes, businesses,
municipalities, and irrigation.  Many of the cooperatives that
currently receive PMA power also have received direct loans or
guarantees from RUS.  According to Western officials, these
cooperatives' financial health depends in part on the availability of
low-cost PMA power.  This is of significant interest to the Treasury
because of its need to recoup the balance these PMA customers owe in
RUS' loans or loan guarantees.\1

Under the status quo, the PMAs' revenues are to repay billions of
dollars of the costs associated with joint\2 and nonpower benefits
for purposes such as irrigation\3 and fish and wildlife protection. 
Because such benefits likely would not cease to exist if power
revenues stopped paying for them, other sources of revenues would
have to be located to fund them.  In order to avoid increasing the
federal deficit, one possible means of paying for these benefits
would be for the Congress to fund them from increased tax receipts. 
However, if federal taxes and revenues could not be increased, then
the Congress would need to offset the spending increase for the
benefits by decreasing federal spending for other purposes. 
Alternatively, some costs could be allocated to categories that are
not reimbursable through power rates or user fees--to flood control
at the Pick-Sloan Program, for example.  However, in such a case,
additional revenues (such as new taxes or new user fees) would be
needed to pay for the costs or offsetting budget cuts to avoid
increasing the budget deficit.  In these cases, because of the need
to find new revenues, uncertainty about repayment of the full
Treasury investment would increase. 


--------------------
\1 GAO/AIMD-97-110 and Rural Development:  Financial Condition of the
Rural Utilities Service's Loan Portfolio (GAO/RCED-97-82, Apr.  11,
1997) discuss the federal government's risk associated with RUS'
borrowers.  In fiscal year 1995, for example, over 150 RUS' borrowers
were preference customers of the three PMAs. 

\2 Joint costs are costs associated with facilities that serve
several purposes.  For example, the dam impounds water not only for
hydropower, but for other purposes of the water project--for
instance, irrigation and recreation. 

\3 For example, we reported that Western was responsible for repaying
about $1.6 billion in irrigation-related costs from power revenues. 
See Federal Electric Power:  Operating and Financial Status of DOE's
Power Marketing Administrations (GAO/RCED/AIMD-96-9FS, Oct.  13,
1995). 


   MAINTAIN FEDERAL OWNERSHIP BUT
   IMPROVE THE MANAGEMENT OF THE
   POWER PROGRAM
---------------------------------------------------------- Chapter 3:2

Many options exist for improving the operations of the hydropower
program while continuing federal ownership.  These options can be
grouped in several different ways, including (1) improving the
planning, budgeting, and funding for capital repairs of the federal
hydropower assets; (2) changing the PMAs' power rates and repayment
methodologies; (3) organizationally restructuring the federal
hydropower program to improve its operating efficiency; and (4)
eliminating the application of selected legal and administrative
requirements to the federal program.  In addition, the government
could dispose of its high-cost hydropower projects.  Some changes can
be made by the PMAs and the operating agencies themselves, while
others would require congressional action. 

Improving the operating efficiency of the federal hydropower program
would not fully respond to the concerns of the advocates of complete
divestiture or privatization, who believe that the government should
not participate in a commercial activity.  Those concerns could be
satisfied only if the hydropower assets were fully divested; however,
improving their operations under federal ownership would better
safeguard the federal investment while continuing to balance the
existing multiple purposes of the projects.  Adoption of these
improvements may have immediate benefits or may be considered an
interim step toward full divestiture, if the Congress proceeds with
that option. 


      IMPROVING CAPITAL PLANNING,
      BUDGETING, AND FUNDING FOR
      REPAIRING THE FEDERAL
      HYDROPOWER ASSETS
-------------------------------------------------------- Chapter 3:2.1

Federal agencies are traditionally funded through annual
appropriations from the Congress.  However, as stated in chapter 2,
the federal budget process does not lend itself effectively to
commercial activities.  Under the current planning and budgeting
process, the Bureau's and the Corps' project and field locations
estimate the costs of and develop the budget proposals for capital
repairs of not only hydropower facilities, but also dams, irrigation
systems, navigation systems, and recreational facilities.  Hydropower
repairs may be assigned lower priorities than other items, and budget
requests are also subjected to 10-percent to 15-percent reduction
targets to reduce the federal deficit.  Under these conditions, the
PMAs' power customers believe, and our previous work showed, that
funding for needed repairs is at best uncertain and at times is not
available when it is needed. 

Several alternatives present themselves for better ensuring that the
federal hydropower resources are repaired in a timely fashion. 
Capital planning and budgeting could be instituted for the federal
hydropower program.  If the PMAs and the operating agencies were to
adopt more businesslike capital planning and budgeting practices,
they would be better able to systematically identify and fund
improvements and repairs to their power systems.  In addition to
capital planning and budgeting, other approaches have been adopted. 
For instance, PMAs, operating agencies, and preference customers have
reached agreements allowing customers to finance some capital
repairs. 


         INSTITUTE CAPITAL
         PLANNING AND BUDGETING
------------------------------------------------------ Chapter 3:2.1.1

The Bureau and the Corps need to improve their planning and budgeting
process to facilitate timely repairs of their hydropower facilities. 
The Corps' need was illustrated in our 1996 testimony on reliability
issues at the Corps' hydropower plants in the Southeast.\4 The Corps
recognized that long-term, comprehensive planning and budgeting
systems are needed to identify and fund key repairs and
rehabilitations at its hydroelectric power plants, especially in the
current environment of static or declining budgets; however, under
its current planning and budgeting system, its funding decisions
cannot be based on such processes. 

Operating under the federal budgeting process,\5 the Corps finds
itself unable to ensure a predictable source of funding for capital
projects at a time when its budget has been decreasing.  Therefore,
it gives priority to routine, ongoing maintenance and performs
reactive, short-term repairs when its power plants experience
unplanned outages.\6 The federal budgeting process does not lend
itself to funding extensive repairs and rehabilitations; when these
actions eventually become essential, the Corps' budgeting process
requires extensive justifications that can take a year or longer to
complete. 

During the early 1990s, the Corps was beginning to address its
planning and budgeting needs, for instance, by beginning to rank
proposed repair and rehabilitation projects.  This effort was
suspended in fiscal year 1995, but the Corps' responsible
headquarters official planned to direct the field locations to
undertake the effort in time to be considered for the fiscal year
1998 budget.  Moreover, in recognition of the need to spend more to
repair and rehabilitate its hydropower plants, the Corps in fiscal
years 1993 through 1997 requested appropriations for major
rehabilitations of some of its hydropower plants.  Ten major
rehabilitation projects have been approved for funding during fiscal
years 1993 to 2007, with a total cost of about $450 million.  These
projects are being funded from the Corps' Construction-General
account generally over a multiyear period and do not need to be
re-budgeted annually. 

As described by Bureau officials, the Bureau's planning and budgeting
process, like the Corps', is lengthy and complex, taking over 2 years
to produce a known budget level.  Because 10-percent to 15-percent
budget cuts are applied to the initial budget and subsequent
proposals made by the regions and their area offices, future funding
levels are uncertain.  For example, Bureau officials in the agency's
Billings, Montana, regional office, described the lengthy budget
process they expected to undergo to achieve a budget for fiscal year
2000.  From the regional perspective, the process began in August
1997 when the regional office received the initial budget proposals
from its area offices.  During the ensuing 16 months, scheduled to
end in December 1998, the area offices, the region, the Bureau's
Denver Office, the Bureau's Washington Office, the Office of the
Secretary of the Interior, and the Office of Management and Budget
will review, discuss, and repeatedly revise the proposed area office
and regional office budgets, resulting in a consolidated budget for
the Bureau and the Department of the Interior.  Although by December
1998 the Department will have informed the regional office of
expected funding levels for fiscal year 2000, certainty about
expected funding levels will not be attained until some time between
February 1999, when the Office of Management and Budget will assemble
and convey the President's budget to the Congress, and October 1,
1999, the start of fiscal year 2000. 


--------------------
\4 GAO/T-RCED-96-180. 

\5 Capital budgeting for federal agencies is discussed in
GAO/AIMD-97-5 and Budget Issues:  Incorporating an Investment
Component in the Federal Budget (GAO/AIMD-94-40, Nov.  9, 1993). 

\6 Problems in funding the maintenance of federal agencies' assets
are discussed in Deferred Maintenance:  Reporting Requirements and
Identified Issues (GAO/AIMD-97-103R, May 23, 1997) and Deferred
Maintenance Reporting:  Challenges to Implementation (GAO/AIMD-98-42,
Jan.  30, 1998). 


         IMPLEMENT ALTERNATIVE
         FORMS OF FINANCING
------------------------------------------------------ Chapter 3:2.1.2

Funding from sources other than federal appropriations has been
suggested as one option to improve how the PMAs and the operating
agencies pay for repairs of the federal hydropower assets.  Although
use of nonfederal funds to finance federal agencies' operations is
generally prohibited unless specifically authorized by the
Congress,\7 several forms of alternative financing have been
authorized by the Congress, according to agency officials. 

Through one type of authorized arrangement, referred to, among other
names, as "advance of funds," nonfederal entities, such as preference
customers, pay for repairs and upgrades of the federal hydropower
facilities.  Under federal budget statutes, such funding must be
ensured before work on a project can be started.  For example,
Western's customers are providing advance funding to renovate the
generating units at the Bureau's Shasta power plant in the Central
Valley Project.  Under an agreement between the Bureau, Western, and
the preference customers, the customers may finance up to $21 million
and deposit the funds in an escrow account to pay for the work.\8 The
Bureau accepts the customers' funds under the Contributed Funds
Act.\9

Customers may be repaid in various ways, including offsets to power
rates under which (1) expenses funded from advances from customers
are excluded from the revenue requirement for repayment purposes or
(2) customers' monthly power bills are credited for the amount each
customer paid to the escrow account.  In the case of the Shasta power
plant, the customers who contributed funds will be issued credits on
their monthly power bills from Western; those that did not contribute
funds will not be issued credits.  According to the Bureau, this
arrangement ensures that all customers contribute.  When completed,
the entire repair cost will have been expensed throughout the
construction period with advance funding from PMA customers. 

Under another form of alternative financing, referred to as "net
billing," invoice amounts are netted out among parties who perform
work or provide services for each other, resulting in the issuance of
one check instead of multiple checks.  Net billing has been used for
purchased power and wheeling for several projects--Central Valley,
Loveland Area, and Pick-Sloan, according to Western officials.\10
Western estimates that the use of net billing has reduced
appropriation requirements by between $40 million to $50 million
annually. 

Under a variation of net billing, referred to as "bill crediting," a
customer agrees to pay one or more of the PMA's bills in exchange for
an equivalent credit on the customer's power bill.  Bill crediting
has the same uses as net billing.  Western estimates that bill
crediting has reduced appropriations' requirements by between $45
million to $60 million annually, mostly in the Central Valley
Project,\11 and that increased use for the Loveland and Pick-Sloan
projects could reduce the appropriations' requirements by between an
additional $2 million to $7 million annually. 

Supporters of alternative financing, among them officials from the
Bureau, the Corps, the PMAs, and the PMAs' customers, note that its
use allows repairs and improvements to be made more expeditiously and
predictably than through the federal appropriations process.  They
believe that alternative financing could provide more certainty in
funding repairs and help address problems such as deferred
maintenance at Corps-operated plants that provide power marketed by
Southeastern.\12 Alternative financing would also move certain costs
out of the budget cycle, decreasing the need for appropriations that
must be repaid through the PMAs' power revenues.  For example, as of
January 1998, Bonneville had entered into long-term agreements with
the Bureau and the Corps that will allow Bonneville to directly fund
about $150 million dollars in capital improvements and operations and
maintenance of the federal hydropower assets in the Pacific
Northwest.  According to Bonneville, these arrangements will shorten
the time needed to secure funding for repairs and maintenance and
will remove maintenance as a funding item that must compete with
other federal budget priorities.  The agreements also promote
coordination between Bonneville, the Bureau, and the Corps in
budgeting for future maintenance and repairs.  Bonneville estimates
that this closer coordination will produce operating efficiencies
that can reduce costs by up to about $48 million per year. 

However, Corps and DOE officials cautioned that expanded use of
alternative financing may not be prudent because, depending on how it
is implemented, oversight by the Congress and the Office of
Management and Budget may decrease.  According to Bureau and DOE
officials, the Congress could take action to foster oversight by the
Congress and other entities.  For example, Bureau officials believe
that to provide for oversight, the agencies could be required to
submit data on expenditures to the Office of Management and Budget
and to the Congress. 

Expanded use of alternative financing may require legislative action,
especially for the projects operated by the Army's Corps of
Engineers.  In a July 1996 memorandum, the Army's Office of the
General Counsel concluded that although the Army has some existing
authority to accept funds from outside parties to finance
replacements, improvements, and other work at the Corps' hydropower
facilities, the use of these funds must be reviewed case by case and
is limited to funds from states and their subdivisions.  According to
the memorandum, the Congress may have to enact more specific
legislation to (1) clarify the terms under which such funds may be
accepted, including the kind of work that they could pay for, and (2)
establish the framework under which the Army, the PMAs, and the
customers should proceed with such alternative financing. 


--------------------
\7 Agencies' use of funds from outside sources without specific
authority is referred to as "augmentation of funds" and is
prohibited. 

\8 According to DOE, the Bureau awarded a contract for the Shasta
rewind project for about $12.2 million in January 1997. 

\9 The Corps' projects are not covered under the Contributed Funds
Act.  However, similar contributions can be now accepted to fund
repairs to the Corps' assets with certain restrictions, according to
the Army's General Counsel. 

\10 Net billing is used pursuant to direction in House, Senate, and
Conference Reports of the 84th Congress and the 1961 Public Works
Appropriation Hearing, according to Western officials. 

\11 Bill crediting is used pursuant to such legislation as the
Reclamation Project Act of 1939 and the Act of August 26, 1937,
according to Western officials. 

\12 GAO/T-RCED-96-180. 


         ESTABLISH ADDITIONAL
         REVOLVING FUNDS
------------------------------------------------------ Chapter 3:2.1.3

The Congress could expand the use of revolving funds.  Under one
revolving fund arrangement, a fund established by a one-time
permanent appropriation is replenished through revenues, which, in
the case of the PMAs, are generated by the sale of power or other
services and credited directly to the fund, instead of being
replenished through annual appropriations.  The Congress has
authorized the use of these funds at such projects as the Colorado
River Storage and Fort Peck projects to fund operation, maintenance,
and replacement costs.\13

Proponents of revolving funds, including some officials of Western,
the Bureau, and a PMA customer group, note that the funds allow
repairs and improvements to be financed more expeditiously and
predictably than the federal appropriations process does.  Like
alternative financing, revolving funds remove some costs from the
budget cycle, thereby decreasing the need for reimbursable
appropriations.  Thus, revolving funds enable the federal
power-related operations to be self-financing and also offer
customers more opportunities to consult with the agencies on how to
spend funds to repair and maintain the hydropower assets. 

However, officials of PMA customer groups and the Office of
Management and Budget also stated that the use of revolving funds
could reduce oversight by external parties such as the Congress and
the Office of Management and Budget and/or may allow repayment
obligations to be incurred that are not routinely approved by these
entities.\14 However, the Congress could be kept informed of the
operating agencies' and the PMAs' spending plans through the annual
appropriations process.  For example, the PMAs could be required to
submit their annual operations and maintenance budgets to the
congressional oversight committees.  A 1993 DOE legislative proposal,
which was not enacted, would have provided for separate accounts
established in the U.S.  Treasury to be funded from all sources,
including sales of power and other services as well as other
collections by, contributions to, and appropriations for
Southeastern, Southwestern, and Western.  These PMAs, the Bureau, and
the Corps would use these accounts to pay for the operations,
maintenance, and rehabilitation of their power assets.  The PMAs
would have submitted their annual operations and maintenance (O&M)
budgets to their budget committees, including estimates of the PMAs'
and the operating agencies' O&M spending, project by project. 
Officials of the Bureau, Western, and a PMA customer group voiced
concerns that revolving funds increase the likelihood that nonpower
costs, such as environmental initiatives and repayment of obligations
to Native Americans, will be added to the revenue requirements base,
with rate impacts that are not fully apparent until later.  For
example, under bills proposed in both the House and the Senate, a
potential future cost of up to about $4.5 million would be financed
with payments from the Upper Colorado River Basin Fund to divest the
lands, structures (including homes), and community infrastructure of
the Bureau's Dutch John, Utah, community that the Secretaries of
Agriculture and of the Interior identify as unnecessary.\15 A Bureau
official estimated that the agency may incur an additional $300,000
over a 2-year period to administer the transfer of assets. 

In a related option, the Congress could authorize the three PMAs to
use a portion of their revenues from power sales to directly fund
statutorily defined hydropower-related activities of the operating
agencies instead of turning the revenues over to the Treasury.  The
Energy Policy Act of 1992, for example, authorizes Bonneville to
directly fund such activities at Bureau and Corps' hydropower
projects in the Pacific Northwest.  If the Congress authorizes other
PMAs to directly fund hydropower assets of their operating agencies,
the PMAs' access to nonappropriated funds, such as those provided to
Bonneville, would be one way to pay for the projects.  The Congress,
however, may wish to consider limiting the types of projects that may
be so funded, as it did for Bonneville. 


--------------------
\13 In addition, since 1974, Bonneville has operated without annual
appropriations by using an agencywide revolving fund maintained by
the Treasury and permanent Treasury borrowing authority.  See
GAO/RCED/AIMD-96-9FS and Bonneville Power Administration:  Borrowing
Practices and Financial Condition (GAO/AIMD-94-67BR, Apr.  19, 1994). 

\14 When it creates revolving funds, the Congress defines the way in
which the funds are used, or it can amend the authorizing legislation
for existing funds to cover additional uses for the funds. 

\15 The community housed Bureau workers while the Flaming Gorge Dam
was built. 


      CHANGE THE PMAS' POWER RATES
      AND REPAYMENT METHODOLOGIES
-------------------------------------------------------- Chapter 3:2.2

Arguments can be made that the way the PMAs establish their revenue
requirements and the way they set their rates need to be changed.  As
noted in our recent products,\16 for example, although generally
following applicable laws and regulations, the PMAs' power rates are
not recovering all of the costs associated with generating,
transmitting, and marketing federal power.  Such cost recovery is
generally required by the Reclamation Project Act of 1939 and the
Flood Control Act of 1944.  DOE's cost recovery order (Order RA
6120.2), however, excludes certain costs associated with facilities
that are not operational and is not specific about the recovery of
other costs.  The PMAs have consequently interpreted the order to
exclude certain costs from their rates.  In addition, the
nonrepayment of some federal investments in hydropower capacity and
other assets (most importantly, irrigation facilities) assigned to
power for repayment raises the issue of whether these investments
will be recovered under the current repayment methods.  In addition,
a question arises about whether the PMAs should be required to
continue to market their power on the basis of cost-of-service
pricing when other parts of the industry are being encouraged to
market their wholesale power on a competitive basis. 

This section discusses various ways that the PMAs could better
recover the costs associated with the federal power program: 

  -- Increasing PMAs' power rates. 

  -- Charging rates based on competition. 

  -- Changing the repayment methodology to recover the federal
     investment faster and decrease the risk of nonrepayment. 

  -- Reallocating costs among the water projects' multiple purposes. 

  -- Merging rate-setting systems to promote the repayment of costs
     at certain facilities. 

Although these changes would address some unrecovered costs that we
identified, they would not address all such costs.  For example, such
unrecovered costs as those associated with the incomplete irrigation
facilities at the Pick-Sloan Program, facilities that are not
operating because of a lawsuit at the Russell project, or
environmental mitigation costs legally exempted from Western's rates
at the Glen Canyon and Shasta dams would not be addressed. 

Several of the methods listed could result in rate increases, but
decisionmakers should consider that increasing the PMAs' rates is in
the government's interest only as long as the rates do not rise to
the point of being noncompetitive.  Because the PMAs already sell
power generated at a few of over 100 federal water projects whose
power they market at prices at or near the prevailing market price, a
rate increase could be counterproductive in these instances\17 and
could not be sustained in a competitive marketplace.  In addition,
some are concerned that rate increases would harm rural communities
and customers. 


--------------------
\16 Power Marketing Administrations:  Cost Recovery, Financing, and
Comparison to Nonfederal Utilities (GAO/AIMD-96-145, Sept.  19, 1996)
and GAO/AIMD-97-110. 

\17 Projects that generate power priced near or above the market rate
or that face competition from other providers include the Central
Valley Project and the Colorado River Storage Project, according to
PMA and PMA customer association officials. 


         INCREASE RATES TO BETTER
         RECOVER ALL COSTS
------------------------------------------------------ Chapter 3:2.2.1

Relying on Office of Management and Budget Circular A-25 on user fees
as well as industry practices and federal accounting standards, our
past reports identified a number of power-related costs that had not
yet been fully recovered through the PMAs' electricity rates.  Such
costs include those for postretirement health benefits and a portion
of Civil Service Retirement System benefits for current employees of
the PMAs and the operating agencies, construction costs for some
projects that were completed or under construction, and construction
and O&M costs for hydropower facilities and water storage reservoirs
that are infeasible and therefore not expected to be completed.\18

Rates could be increased to fully recover some of these costs.  For
instance, the full costs associated with the postretirement health
benefits and the Civil Service Retirement System benefits could be
recovered through power rates.  The three PMAs will begin the process
of recovering pension and postretirement health benefit costs by
including the unfunded liability of the Civil Service Retirement
System and postretirement health and life insurance costs of
power-related employees in their power repayment studies, beginning
in fiscal year 1998.\19

Revenues from rate increases could also pay for unrecovered capital
costs for projects that are under construction or not yet in
commercial operation when those projects are brought on line.  Under
DOE's repayment guidance, the recovery of some federal investments in
hydropower has been deferred until projects are completed and placed
into commercial operation.  These costs are to be repaid when these
projects come on line, although rate increases may be substantial. 
For example, a Southeastern official stated that the costs for the
nonoperational pumping units at the Corps' Russell project, which he
estimated at about $528 million as of August 1997, are not yet
subject to repayment.  Because of litigation over large fish kills,
these units have not been allowed to operate commercially and these
costs have not been included in Southeastern's rates.  However, if
the nonoperational units come on line, these costs would be recovered
through rates.  The resulting rate increase for customers of that
particular rate-setting system may be as high as 25 percent, but in
this instance the power would still be competitively priced,
according to this official. 


--------------------
\18 GAO/AIMD-97-110; Federal Electricity Activities:  Appendixes to
the Federal Government's Net Cost and Potential for Future Losses,
Volume 2 (GAO/AIMD-97-110A, Sept.  19, 1997); GAO/AIMD-96-145; and
GAO/T-RCED-96-142. 

\19 Bonneville plans to begin recovering these costs in fiscal year
1998; full recovery is planned beginning in fiscal year 2002. 
Consistent with current policies and law, the PMAs do not plan to
recover pre-fiscal year 1998 net costs. 


         CHARGE RATES BASED ON
         COMPETITION
------------------------------------------------------ Chapter 3:2.2.2

The industry is being encouraged to base its power rates on a
competitive basis rather than on cost of service.  Therefore, the
Congress could enact legislation authorizing or directing the PMAs to
change from cost-of-service rates to rates based on competition.\20

In accordance with legislation, the PMAs are to set their rates at
the lowest possible level consistent with sound business principles
and market their power primarily to preference customers.  Because
the three PMAs' overall average revenue per kWh is at least 40
percent below existing market rates,\21 charging market rates for PMA
power would most likely cause the PMAs' rates to rise.\ \22 With
higher rates, the PMAs' revenue would be likely to increase and,
consequently, the risk of nonrepayment of the federal investment
would be likely to decrease as long as the rates remain competitive
relative to prevailing market rates. 


--------------------
\20 Except as otherwise provided by law, the PMAs are directed to
generate revenues sufficient to recover all costs incurred as a
result of generating, transmitting, and marketing electric power,
including repayment of the federal investment and other debt with
interest.  In addition, legislation makes Bonneville and Western
responsible for repaying, through power revenues, some irrigation
costs associated with the hydropower projects.  DOE requires each PMA
to annually prepare a repayment study to test the adequacy of its
rates and to show, among other things, estimated revenues and
expenses, estimated payments on the federal investment, and the total
amount of federal investment to be repaid. 

\21 GAO/RCED/AIMD-96-9FS; GAO/AIMD-96-145; and GAO/AIMD-97-110 and
110A. 

\22 As noted earlier, rates for a few projects are already at or near
the market price.  Also, prices above the market rate could not be
sustained in a competitive marketplace. 


         CHANGE THE REPAYMENT
         METHODOLOGY TO RECOVER
         FEDERAL INVESTMENT MORE
         QUICKLY
------------------------------------------------------ Chapter 3:2.2.3

The Congress or the Secretary of Energy could require the methodology
for repaying PMA debt to be changed in order to recover the federal
investment more quickly.  Such a change could increase the PMAs'
rates and revenues as well as the rate of repayment to the Treasury. 
Under DOE's current policy and consistent with applicable laws, the
PMAs may defer repayment of annual expenses when power revenues do
not meet repayment needs during low water years.  Deferred annual
expenses accrue interest at a current interest rate until they are
repaid and generally must be repaid prior to the PMAs' repaying the
principal investment.  When repaying principal investment, the PMAs
generally must repay their highest interest-bearing debt first rather
than the oldest debt.\23 These provisions establish some of the
financing flexibility the PMAs need because their revenue reflects
the year-to-year variability of water flows and hydropower
generation; however, they also result in rates that are lower than
they otherwise would be, slower repayment of the federal investment,
and a net cost to the Treasury because interest rates on the
outstanding federal investment are substantially below the rates
Treasury incurs to provide funding to the PMAs and other federal
programs.  Repaying the federal investment faster would decrease the
Treasury's interest costs and the amount at risk for nonrepayment. 
However, as for any alternative that increases rates, policymakers
would need to consider the impact on the PMAs' customers and their
region. 


--------------------
\23 Policies Governing Bonneville Power Administration's Repayment of
Federal Investment Still Need Revision (GAO/RCED-84-25, Oct.  26,
1983); GAO/RCED/AIMD-96-9FS; GAO/AIMD-96-145; and GAO/AIMD-97-110. 


         REALLOCATE COSTS AMONG
         THE PROJECTS' MULTIPLE
         PURPOSES
------------------------------------------------------ Chapter 3:2.2.4

The Congress, or in some cases the operating agencies, could revise
the formulas used to allocate costs currently assigned to the
multiple purposes of the federal water projects or the "joint costs"
(those shared among more than one of the purposes--for example, the
capital costs associated with the dam).  In some cases, this action
would reduce the capital investment that would have to be repaid
through the rates the PMAs charge for electricity.  For example,
officials of the Corps and Western's preference customers noted that
some projects currently allocate little or no costs to recreation or
water quality, even though these categories have become increasingly
important purposes since the operating agencies prepared the project
cost allocations.  Through reallocation, a portion of the costs
assigned to power would be reassigned to recreation and the electric
rates could be lowered accordingly. 

However, reallocations could result in some costs that are currently
being repaid through power revenues--for example, most
irrigation-related costs--needing to be repaid through other
means.\24 Absent action by the Congress or the operating agencies to
institute or increase existing user fees for the activities currently
repaid through power revenues, these costs could end up not being
repaid.  Thus, while the PMAs' ratepayers could be relieved of the
repayment burden of costs no longer assigned to power, the federal
taxpayer may end up bearing the burden instead.  Also, in commenting
on our draft report, DOE's Power Marketing Liaison Office noted that
the equity of certain project beneficiaries (for example, power
customers) having to repay more than their fair share of multipurpose
costs also needs to be addressed. 

In some cases, congressional action would be required to authorize a
reallocation of costs.  For example, as of September 30, 1994, the
federal government had about $454 million in federal investment (1)
in the Pick-Sloan Program's hydropower capacity that was initially
designed to be used by future irrigation projects and (2) in the
costs associated with storing water for these projects.\25 Although
these costs are scheduled to be repaid through Western's power
revenues, under Western's statutory repayment principles, these
costs, which we estimated at $464 million as of September 30, 1996,
cannot be recovered unless the associated irrigation projects come
into service.  According to the Bureau, however, almost all of these
planned irrigation projects are infeasible and are unlikely to be
completed.  Reallocating the $464 million from irrigation to
hydropower would help ensure full recovery, but without legislative
action to do so, it is probable that Western's power rates will not
recover the principal or any interest on it. 


--------------------
\24 As of September 30, 1994, the Secretary of the Interior had
assigned over $2.4 billion in irrigation-related costs to Bonneville
and Western for repayment through power revenues. 

\25 GAO/T-RCED-96-142. 


         MERGE RATE-SETTING
         SYSTEMS
------------------------------------------------------ Chapter 3:2.2.5

For some facilities, rate-setting systems could be merged to expedite
repayment.  For example, at two facilities--the Stampede Powerplant
at the Bureau's Washoe Project and the Mead-Phoenix Transmission
Line, which is partially owned by Western, with a combined federal
investment of at least $108 million, as of September 30,
1996--Western generated insufficient income to recover capital and
operating costs.  Western officials are considering a merger of the
Washoe and Mead-Phoenix rates with others, resulting in blended rates
and increasing somewhat the likelihood of full repayment of the
federal investment.\26


--------------------
\26 GAO/AIMD-97-110A. 


      RESTRUCTURE THE FEDERAL
      HYDROPOWER PROGRAM TO
      IMPROVE OPERATING EFFICIENCY
-------------------------------------------------------- Chapter 3:2.3

In recognition of the changing power markets, the Congress could
restructure the PMAs organizationally to better enable them to
compete.  It can be argued that such changes could provide the PMAs
with the flexibility to respond better to market changes and to the
needs of their customers, thereby helping to ensure the PMAs'
survival and the repayment of the federal investment.  It can also be
argued that the PMAs' federal responsibilities should be continued
because of the need to balance the multiple purposes of the water
projects.  Also, restructuring the PMAs may be seen as an interim
step to privatizing them and the operating agencies'
hydropower-related assets. 

However, absent congressional action and depending on how the program
might be reorganized, any restructuring of the PMAs that increases
their operational independence may decrease congressional and other
oversight.  TVA, a wholly owned federal utility with little external
oversight, used its financial ties to the federal government and its
operational independence to embark on an ambitious nuclear power
building program that resulted in nearly $28 billion in debt, as of
September 30, 1996.  This debt puts TVA at a competitive
disadvantage, especially if the Congress were to revise legislation
and require TVA to compete with other power suppliers.  TVA's
experience highlights the need for the Congress to carefully consider
what oversight would be needed before allowing the PMAs to
restructure to be more competitive. 


         REORGANIZE THE PMAS AS
         FEDERALLY OWNED
         CORPORATIONS
------------------------------------------------------ Chapter 3:2.3.1

The Congress could enact laws to authorize the PMAs to operate as
federally owned corporations.  This type of restructuring,
"corporatization," would allow a government entity that serves a
public function of a predominantly business nature to operate in a
more efficient, businesslike fashion, while preserving the public
service goals that are unique to federal agencies (for example,
revenues from Western's sale of power are scheduled to pay for most
of the federal investment in irrigation facilities).\27

Establishing a PMA as a government corporation has been formally
proposed in recent years.  In 1994, a proposal was drafted to
corporatize Bonneville as a way to help maintain its competitiveness. 
Bonneville has been faced with competition from alternative power
sources with lower costs, debt that exceeded $17 billion as of
September 30, 1996, and upward pressure on its costs, caused in part
by expanded, more costly efforts to protect salmon.  The proposal was
based on a recommendation in a National Academy of Public
Administration report that examined alternative structures to achieve
the maximum efficiency and effectiveness at Bonneville.  The
administration considered legislation to make Bonneville a wholly
owned government corporation under the Government Corporation Control
Act.  This action was intended to increase Bonneville's flexibility
over personnel; procurement; property management; and budgetary,
litigation, and claims settlement functions and to enable Bonneville
to compete more effectively in electric power markets.\28 Bonneville
estimated that the savings from corporatization would have been as
much as $30 million annually.  In that the other three PMAs'
operations are much smaller than Bonneville's,\29 the estimated
savings from their corporatization would likely be smaller. 

Corporatization may permit repairs and improvements to be financed
more expeditiously and predictably than the federal appropriations
process.  Presuming that a revolving fund would be established as
part of the corporatization, the corporation could operate in a
businesslike fashion, without having to submit a budget request for
annual appropriations to finance operations.  Although the electric
utility industry is now unbundling its services, depending on how the
government corporation was structured, the generation, transmission,
and marketing aspects could be put under one agency, possibly
reducing overhead.  Each PMA could be established as a separate
corporation or two or more of the PMAs--Southeastern and
Southwestern, for instance, could be merged.  The latter option may
afford the economies of scale necessary to make the new corporation
or corporations viable, according to a Corps headquarters official. 
Alternatively, distinct federal rate-setting systems could be
corporatized as separate entities from the rest of the PMA.  Western
officials responsible for marketing power from the Bureau's power
plants within the Salt Lake City Integrated Projects--the Colorado
River Storage Project plus the Provo River, Falcon-Amistad, and other
projects that are aggregated for rate-setting purposes--suggested
that their marketing program could be corporatized.  They said that
it already benefits from substantial operating and budgeting
independence because its operations are financed from a revolving
fund.  However, in its response to our draft report, DOE's Power
Marketing Liaison Office stated that it is not Western's policy to
support the corporatization of this marketing program at this time. 

If the government's objective is to eventually end its participation
in a "commercial" activity, corporatization could be an interim step
toward divestiture of its hydropower-related assets.  In a 1995
report on the privatization or divestiture practices of other
nations, we noted that the five nations\30 we reviewed generally (1)
converted government agencies or functions into a corporate form
before privatizing them or (2) primarily privatized entities already
in a corporate form.\31 Converting a government department into a
corporate entity, followed in many cases by a privatization, has been
common worldwide during the past decade. 

In New Zealand, for example, the government included a set of reform
principles designed to improve performance in the delivery of public
sector goods and services in the State-Owned Enterprises Act of 1986. 
The government anticipated that entities corporatized under this act
would be subject to the same regulation, antitrust, tax, and company
law as private enterprise.  The restructuring of the electricity
industry commenced with the corporatization of the government's
generation and transmission capacity in 1987, corporatization of the
retail power companies in 1993, full deregulation of the retail
sector in 1993 and 1994, and establishment of a competitive wholesale
electricity market in 1996.  According to a former New Zealand
government official, the government privatized seven small
government-owned generating projects in 1995.  Additional
privatizations of generation facilities, while possible, are not
anticipated, according to New Zealand's Energy and Finance Ministers. 

The changes in electricity rates since the New Zealand's
restructuring of the electricity sector are noteworthy, according to
a former New Zealand government official we interviewed.  Although
very large rate increases had been feared for farmers, for example,
rural rates declined by about 40 percent in real terms from 1987,
when the reform process started, to 1994, according to one study.\32
Cross subsidies between customer classes are reported to be greatly
reduced.  Over a longer term, inflation-adjusted retail domestic
(residential) rates increased by about 5 percent to 15 percent from
1985 through 1997 and from about 16 percent to 20 percent from 1990
through 1997, according to the New Zealand Ministry of Commerce. 
Commercial rates, on the other hand, decreased by about 20 percent to
28 percent from 1985 through 1997 and by about 1 percent to 9 percent
from 1990 through 1997.\33

In the United States, experience with such conversions after interim
corporatization of government activities has been limited.  For
example, the Congress enacted legislation in 1992 to corporatize
DOE's uranium enrichment operations as the U.S.  Enrichment
Corporation\34 in a transitional step toward eventual
privatization.\35 Similarly, a bill now in House committees would
convert the three PMAs into corporations as an interim step toward
their privatization. 

Despite the advantages, creation of a government corporation could
significantly reduce the amount of oversight the entity receives.  In
the past, we have suggested that the Congress strengthen the
oversight and accountability of government corporations.\36 For
example, over the years, we\37 and others, have characterized TVA, an
existing wholly owned federal corporation, as having insufficient
independent oversight.\38 Some have noted, moreover, that an entity
that resulted from a merger of, for instance, the Bureau's water
management and power generating responsibilities with Western's power
marketing responsibility could experience conflicts among these three
different roles. 


--------------------
\27 A corporate form of organization may be appropriate for the
administration of government programs that are predominantly of a
business nature, produce revenue and are potentially self-sustaining,
involve a large number of businesslike transactions with the public,
and require a greater flexibility than the customary type of
appropriations budget ordinarily permits.  See Government
Corporations:  Profiles of Existing Government Corporations
(GAO/GGD-96-14, Dec.  13, 1995). 

\28 The draft bill to corporatize Bonneville contained many specific
statutory and regulatory exemptions, which are described in detail in
Government Corporations:  Profiles of Recent Proposals
(GAO/GGD-95-57FS, Mar.  30, 1995). 

\29 For fiscal year 1995, Bonneville had total operating revenues of
about $2.4 billion compared with about $159 million for Southeastern,
$114 million for Southwestern, and $713 million for Western,
according to the PMAs' annual reports. 

\30 The five nations are Canada, France, Mexico, New Zealand, and the
United Kingdom. 

\31 Budget Issues:  Privatization/Divestiture Practices in Other
Nations (GAO/AIMD-96-23, Dec.  15, 1995).  See also Deficit
Reduction:  Experiences of Other Nations (GAO/AIMD-95-30, Dec.  13,
1994). 

\32 "The Impact of Electricity Reforms on Rural New Zealand," P.  J. 
Farley, 1994. 

\33 Comparable data for industrial rates were not available. 

\34 See Uranium Enrichment:  Observations on the Privatization of the
United States Enrichment Corporation (GAO/T-RCED-95-116, Feb.  24,
1995) and Uranium Enrichment:  Activities Leading to Establishment of
the U.S.  Enrichment Corporation (GAO/RCED-94-227FS, June 27, 1994). 

\35 The corporation was formed in 1993 and its sale was authorized by
the President in July 1997.  However, as of February 1998, the
transfer to private status, which is expected to be completed in
1998, had not been completed. 

\36 Congress Should Consider Revising Basic Corporate Control Laws
(GAO/PAD-83-3, Apr.  6, 1983). 

\37 GAO/AIMD/RCED-95-134; Triennial Assessment of the Tennessee
Valley Authority--Fiscal Years 1980-82 (GAO/RCED-83-123, Apr.  15,
1983); and Tennessee Valley Authority--Options for Oversight
(GAO/EMD-82-54, Mar.  19, 1982). 

\38 A bill has been introduced in a Senate Committee to address this
issue by replacing TVA's current three-member board of directors with
a nine-member board. 


         CONSOLIDATE POWER-RELATED
         FUNCTIONS UNDER ONE
         AGENCY
------------------------------------------------------ Chapter 3:2.3.2

The Congress could consolidate the power-related operations of the
operating agencies and the PMAs.  Some operational improvements and
cost savings could result.  Officials at the Bureau's Denver office
recommended that Western's assets be returned to the Bureau so that
the Bureau could better coordinate the multiple purposes of the water
projects, while reducing overhead.\39 They estimate that overhead
costs could be reduced by up to 30 percent if Western's power
marketing activities were consolidated within the Bureau. 

Although the Bureau and the Corps previously marketed the power they
generate, concerns exist about reconsolidating the power marketing
function in these agencies because of the need to balance the needs
of hydropower with the needs of the other activities the agencies
pursue.  Each agency has it own priorities, which do not always favor
maximizing power revenues.  For example, the Congress may provide
funds to the Corps to upgrade a failing generator, but if a key lock
in the Corps' navigation system were disabled, the Corps might divert
the funds intended for the generator to the lock.  This could prolong
an outage at the power plant and cause the government to lose
revenue.  Although a Corps headquarters official stated that this
scenario occurred infrequently, he said that a repair project may be
deferred because of conflicting priorities.  At the same time, if the
power generating activities of the Corps and the Bureau were
consolidated within the PMAs, the PMAs, which have a primary mission
of marketing power, may inadequately consider the other purposes of
the water projects when operating the power plants.  In addition,
consolidations clash with the developing trend among vertically
integrated power utilities to segregate generation, transmission,
distribution, and ancillary services. 


--------------------
\39 The Bureau owned and operated Western's marketing and
transmission assets before Western's creation in 1977. 


      ELIMINATE SELECTED LEGAL AND
      ADMINISTRATIVE REQUIREMENTS
-------------------------------------------------------- Chapter 3:2.4

Bureau, Corps, and PMA officials believe that some of the legal and
administrative requirements that their agencies must follow cause
them to operate in an unbusinesslike fashion and may cause the PMAs'
power rates to increase.  For example, aware of the need to operate
more efficiently, in February 1996 Western chartered an internal
study designed to identify and address laws, regulations, and rules
that it determined to be counterproductive to its functioning in a
businesslike manner.  Although many of the study's recommendations
are administrative in nature, Western identified opportunities to
improve its performance that ranged from a few thousand dollars to
millions of dollars.  For example, the report on the study recommends
that Western request an exemption from DOE's requirement to report
quarterly on safety.  Western contends the report is of no value, but
exempting it from this requirement could save Western $6,630
annually.  In another example, Western estimated that if it used a
credit card to purchase supplies and services instead of purchase
orders, it could save over $500,000 annually.  In an example that
would require legislative action, exempting Western from the
statutory requirements in the Federal Acquisition Regulations about
taking sealed bids for procurements could save the agency $115,600
annually.  Of more consequence, the Congress could allow Western to
pay prevailing local area wages instead of those required by the
Service Contract Act of 1965.  The report states that such an
amendment could save Western about $6.2 million annually.  The scope
of Western's study included the Code of Federal Regulations, the
Federal Acquisition Regulations, executive orders, DOE's orders and
guidelines, and other directives.\40


--------------------
\40 In January 1998, a draft report was released for public comment
that recommended $159 million in cost savings for Bonneville, which
included at least $10 million annually from legislative changes in
procurement and personnel laws designed to improve administrative
effectiveness and efficiency. 


      DISPOSE OF HIGH-COST
      HYDROPOWER PROJECTS
-------------------------------------------------------- Chapter 3:2.5

The Congress could pass legislation that would allow the Bureau and
the Corps to divest themselves of projects that have power generating
costs that exceed the costs and rates of their rate-setting system. 
Officials from the Bureau, officials from two of Western's customer
groups, and representatives of some of Southwestern's customers
suggested that the PMAs could operate more efficiently and reduce
pressure to raise power rates if the operating agencies were allowed
to dispose of several plants that produce higher-cost power.\41
Collectively, they suggested that some of the hydropower plants at
the Bureau's Collbran, Dolores, Loveland Area, and Rio Grande
projects as candidates for disposal.  According to Bureau officials,
some of these projects associated with the Colorado River Storage
Project produce power at costs ranging from about 3.5 to 6 cents per
kWh, whereas Western sells power at a composite firm rate of about 2
cents per kWh for the Colorado River Storage Project.  According to a
Corps official, one obvious problem with this option is finding a
willing buyer for these inefficient units.  Also, to the extent that
power revenues cease to pay for some of the federal investment in
constructing these units, the taxpayers would assume a larger burden. 
Whether the government's investment in these projects is fully
recovered depends on the terms and conditions of the sale and the
resulting price received for the assets. 


--------------------
\41 According to the Bureau, its divestment policy suggests that it
can only divest isolated or remote water projects that do not have
international or interstate ramifications. 


   DIVEST THE FEDERAL HYDROPOWER
   ASSETS
---------------------------------------------------------- Chapter 3:3

Consistent with the philosophy that the government should not be
involved in commercial activities that are best left to the
nonfederal or private sector, the Congress could enact legislation to
divest the PMAs and the government's hydropower assets.  As we
concluded in our March 1997 report, divesting the federal hydropower
assets, while possible, would be complicated for several reasons.\42
Any divestiture of hydropower-related assets would need to balance
the multiple purposes of the water projects that limit and define how
water is released through the turbines, how and when electricity can
be generated, and in what quantities.  These federal responsibilities
would not necessarily terminate after a divestiture.  Other factors
would also have to be accommodated.  These factors include the types
of assets being divested, the conditions attached to the sale and the
use of the assets after the divestiture, the operating conditions of
the assets, the sales mechanism used, and the impact of the
divestiture on regional economies, including the impact on regional
electricity prices.  Of particular note, the impact of a divestiture
on the future rates of the preference customers would have to be
considered.  If the PMAs were privatized, rates would likely increase
to varying degrees for most of the current preference customers. 
Together, these factors complicate the sale of federal hydropower
assets and at the same time could affect the willingness of potential
buyers to bid on the federal hydropower assets and the price the
government could obtain for them.  It should be noted that customers
themselves have proposed defederalization of the federal hydropower
assets.  For example, in 1995, 37 of Western's preference customers
advocated an arrangement whereby they would purchase, lease, or
obtain other rights to the federal hydropower generating assets
within the Boulder Canyon and Parker-Davis projects, as well as
certain transmission projects.  According to a representative of
these customers, this proposal was made to prevent an investor-owned
utility from acquiring the federal power resources and was also a
reaction against other privatization proposals that were being
presented at that time. 


--------------------
\42 GAO/RCED-97-48. 


      ACCOMMODATING MULTIPLE
      PURPOSES AND OTHER PUBLIC
      POLICY FACTORS
-------------------------------------------------------- Chapter 3:3.1

With very few exceptions, federal hydropower projects have multiple
purposes specified in their authorizing legislation.  For example,
the Corps' Fort Peck project on the Missouri River in Montana has
hydropower as a purpose as well as providing for fish and wildlife
habitat, flood control, irrigation, navigation, recreation, water
quality, and water supply.  Multiple purposes are often complementary
but are sometimes at odds.  For example, water is stored in and
released from a reservoir to provide for recreation, but its release
through the turbines could be scheduled in a way that is intended to
maximize revenue.  In contrast, Western's Billings, Montana, office
forecasts decreases in power revenues in the long-term because water,
which would otherwise be used to generate electricity, will be
increasingly used for irrigation and for other purposes.  In its
fiscal year 1995 repayment study, Western predicted that revenues
from the sale of hydropower would decrease from about $253 million in
2001 to about $213 million (in constant 1995 dollars) in fiscal year
2080 for the Pick-Sloan Program. 

At the Bureau's and the Corps' water projects, power generation is
defined and constrained by the requirement to manage the water for
other purposes.  The Bureau, for instance, at some projects has
restricted releases through the turbines to mitigate environmental
impacts downstream.  The need to manage water for multiple purposes
and to generate hydropower in a way that balances other purposes
would have to be accommodated even after a divestiture occurs, absent
congressional action. 

In addition, the water rights of Native Americans and of states would
need to be accommodated in the event of a divestiture.  According to
Bureau officials, Native Americans' rights to water at some federal
water projects are the earliest and thus supersede the use of water
for other purposes, including hydropower generation.  As an example,
Bureau officials cited a legal settlement with tribal entities of the
Fort Peck Reservation in Montana that includes the right to about 1
million acre-feet of water from the Missouri River.\43 In addition,
according to DOE's Power Marketing Liaison Office, a divestiture may
have to address how to transfer out of federal ownership the
transmission lines and rights-of-way that traverse tribal lands.  The
tribes may be concerned about the transfer or sale of such lines to
private parties. 

States also have water rights, and the Bureau and the Corps are
increasingly arbitrating between the claims of various states.  For
example, for several years, Alabama, Florida, and Georgia have been
contesting the uses of water in two river basins in the Southeast
that the Corps manages. 


--------------------
\43 One acre-foot is the amount of water that it would take to cover
1 acre of land with water to a depth of 1 foot. 


      REGULATION OF HYDROPOWER
      ASSETS WOULD BE AFFECTED BY
      THE TYPES OF ASSETS DIVESTED
-------------------------------------------------------- Chapter 3:3.2

As stated in our March 1997 report, the three general ways the
government could divest itself of its hydropower assets are divesting
(1) only the PMAs (including the right to market power and any
associated federally owned transmission assets); (2) the PMAs and the
generating assets of the Bureau or the Corps or both; and (3) the
PMAs, the generating assets, and the balance of the projects (for
example, the dams and the reservoirs).\44 Divesting combinations of
these assets is also possible.  In general, divesting only the PMAs
and the hydropower generating assets would be less complicated than
divesting the balance of the projects because the first two
alternatives retain the Bureau and the Corps in their role of
managing how water is used and in balancing the projects' multiple
purposes.  The kinds of assets divested will influence the regulatory
issues accompanying a divestiture. 

Many options for regulating the operations of divested hydropower
assets exist, including regulatory regimes that could be established
by federal, state, or regional authorities.  FERC, which currently
licenses the operation of nonfederal hydropower assets, primarily
regulates the reasonableness of wholesale rates charged by the PMAs
but does not provide more detailed oversight.  According to FERC
officials, FERC has experience regulating the multipurpose aspects of
water development at over 1,600 projects nationwide pursuant to much
the same multiple-use standards as apply to federal projects.  FERC,
however, does not have complete authority to set regulatory
requirements.  Other federal and state agencies, through FERC's
regulatory process, may impose mandatory conditions on FERC's
licenses, which complicate FERC's licensing process. 

If only the PMAs (including their rights to sell power and any
transmission lines) were divested, then the Bureau and the Corps
would continue to operate the hydropower plants, dams, and reservoirs
in accordance with existing plans, guidelines, and regulations.  In
such a case, the buyer would not need a FERC-issued license; the
Bureau and the Corps would continue to manage the water as in the
past, the existing restrictions would be likely to remain in effect,
and the buyer would market the power subject to the same conditions
as the former PMA.  According to FERC officials, they prefer to
license all of a project's features that have a role in power
production.\45

However, if the power plants were divested as well, the new owner
would be required to obtain an operating license from FERC, unless
this requirement was specifically exempted by law.\46 Licensing a
divested plant could take a long time.  We reported, for example,
that the median processing time for 111 projects applying for
relicensing between January 1982 and May 1992 was 2.5 years.\47 Some
had taken as long as 10 to 15 years.  In January 1998, a FERC
official told us that the median time to relicense over 150 projects
whose licenses expired in 1993--the most recent data FERC had
analyzed--was about 30 months. 

If a divestiture involves a PMA, the power plants, and the balance of
the water projects (most importantly, the dams and reservoirs), the
Bureau and the Corps would no longer fill the role of specifying the
operating conditions of the project.  Instead, safeguards for the
multiple uses of the water would primarily be contained in the
conditions FERC would attach to the operating license pursuant to the
Federal Power Act.  In such an event, in licensing the hydropower
plant, FERC would be required to weigh the plant's impact on such
aspects as the environment and recreation.  Licensing would therefore
be complicated by the need to complete a number of studies on the
power plant's impact on fish, plant, and wildlife species; water use
and quality; and any nearby cultural and archeological resources. 
Moreover, the government of each affected state would have the
opportunity to issue a water quality certification. 

FERC officials also cautioned that if power plants, dams, and
reservoirs were sold, then FERC's licensing process could revisit the
management and uses of the water pursuant to the Federal Power Act
and possibly change the operation of the project, potentially
affecting power generation.  In connection with this issue, the
executive director of the National Hydropower Association stated that
nonfederal hydropower plants are losing generating capacity because
of environmental restrictions or mitigations that are attached as
conditions to their operating licenses as FERC relicenses those
plants.  Moreover, according to a September 1997 report by DOE's
Idaho National Engineering Laboratory,\48 at the time of relicensing,
96 percent of the peaking projects relicensed since 1987 have had
their ability to meet peak demand reduced.  Of the 52 projects that
were relicensed from 1987 to 1996, FERC added capacity to only 4
projects, but the remaining 48 projects had their ability to meet
peak demand reduced by from 0.4 to 54.3 percent of their previous
capacity:  the average reduction was 6 percent.  Also, FERC's review
of over 130 projects licensed from the 157 applications filed in 1991
shows that while generating capacity had a very small increase,
actual electricity generation had a very small decrease--less than 1
percent. 


--------------------
\44 GAO/RCED-97-48. 

\45 See FERC testimony of June 10 and October 7, 1997, before the
Subcommittee on Water and Power, Senate Committee on Energy and
Natural Resources. 

\46 For example, a bill currently before a House committee would
specifically grant the new owner a conditional 10-year license for
continued operation and maintenance of the hydropower facility. 
Thereafter, a FERC license would have to be obtained. 

\47 Electricity Regulation:  Electric Consumers Protection Act's
Effects on Licensing Hydroelectric Dams (GAO/RCED-92-246, Sept.  18,
1992). 

\48 Hydropower Resources at Risk:  The Status of Hydropower
Regulation and Development--1997 (DOE/ID-10603, Sept.  1997). 


      TRADE-OFFS EXIST BETWEEN THE
      CONDITIONS ATTACHED TO THE
      SALE AND USE OF ASSETS AND
      THE BIDS RECEIVED
-------------------------------------------------------- Chapter 3:3.3

The explicit and implicit liabilities borne by the government and
which of those liabilities would transfer to a buyer would also
affect the price obtained for the federal power assets.  Sales of
some or all of the hydropower assets--at prices that exceed the value
to the government--would produce budgetary savings in the long run,
according to a November 1997 report by the Congressional Budget
Office.\49 The report estimates that the combined assets of the three
PMAs may be worth between about $8 billion and $11 billion.  A sale
could also result in a future stream of tax payments to the Treasury,
also depending on the divestiture's terms and conditions.  However,
the report states that losses are possible, depending on the terms
and conditions of the sale.  In addition, as a matter of general
principle, policymakers would need to take into consideration the
fact that assets that are sold with many or relatively onerous
restrictions (from the viewpoint of a prospective purchaser) or
uncertainties about future operations are correspondingly less
attractive and are likely to sell for less.  While the government may
still choose to place restrictions or to assign or retain certain
liabilities, the financial consequences in terms of the sale price
should be assessed. 

If the government's objective is to obtain the maximum possible price
for its assets, the government could retain certain liabilities that
could reduce risks to potential buyers.  In some cases, the federal
government could be in a better position than the buyer to bear
certain risks.  For instance, in the proposed divestiture of the U.S. 
Enrichment Corporation, the government would retain liability for the
environmental cleanup associated with the prior production of
enriched uranium.  According to a contractor's report,
decontamination and decommissioning activities at uranium enrichment
plants could cost as much as $17.4 billion in 1994 constant dollars. 
At some hydropower projects, available generating capacity has been
diminished by up to one-third because of the need to mitigate
environmental impacts downstream.  Buyers may discount any prices
they offer because of the loss of available generating capacity
unless the government assumes the liability for mitigating
environmental impacts.  In addition, in the case of the federal
hydropower assets, uncertainty about future operating conditions
because of potential environmental liabilities may discourage bidding
or result in lower prices than if the federal government assumes some
of the liabilities.  For instance, one provision of the Central
Valley Project Improvement Act directs the Secretary of the Interior
to manage annually 800,000 acre-feet of water for environmental
purposes authorized by the act.\50 According to the Bureau, an
analysis of the environmental impacts indicates that hydropower
generation may be reduced by about 5 percent.  Were the government to
divest the project's assets, it might agree to limit the effect of
water use restrictions on potential buyers for a specific period and
to specify changes in water use restrictions over time to reduce the
uncertainty the buyer would face. 

If the government's objective is to expedite the divestiture on terms
that would less adversely affect the projects' beneficiaries, getting
the highest possible price for the assets might be a secondary
consideration.  For example, although a decision to limit bidders on
particular assets to certain geographic areas would foster a goal of
local or regional control of those assets and expedite a transfer, it
could reduce the proceeds from the sale if other potentially
interested buyers were precluded from making offers.  In the ongoing
divestiture of the Alaska Power Administration, an overriding concern
is to protect that PMA's ratepayers from increases in electricity
rates.  Decisionmakers therefore restricted the eligibility of
bidders to only nonfederal entities from within the state of Alaska. 
It also accepted a sale price approximating the present value of
future principal and interest payments that the Treasury would have
received instead of establishing the price by selling the assets in
an open, more competitive fashion to the highest bidder. 


--------------------
\49 Should the Federal Government Sell Electricity?, Nov.  1997. 

\50 In addition, two other provisions could eventually allocate up to
another 600,000 acre-feet of water for fish and wildlife mitigation
at wetland refuges and the Trinity River, according to estimates by
the Congressional Budget Office. 


      TRADE-OFFS BETWEEN ASSETS'
      OPERATING CONDITIONS AND THE
      NEED TO IMPROVE THEM MUST BE
      CONSIDERED IN THE EVENT OF A
      DIVESTITURE
-------------------------------------------------------- Chapter 3:3.4

Assets that are in better operating condition are more likely to
attract higher bids than assets in poor condition.  We testified in
July 1996 that federal hydropower plants in the Southeast have
experienced significant outages and that these outages occur because
of the age of the plants--an average of about 30 years--and the way
they have been operated.\51 If these hydropower assets were to be
sold without reducing the current backlog of necessary maintenance,
bids would be lower.  However, a 1995 World Bank review of
international experience with divestitures found that in preparing a
government enterprise for divestiture, a government should generally
refrain from making new investments to expand or improve that
enterprise because any increase in sales proceeds is not likely to
exceed the value of those investments.  DOE's Power Marketing Liaison
Office noted that the statement of the World Bank should not be
interpreted to imply that federal facilities should be allowed to
decay without proper maintenance. 


--------------------
\51 GAO/T-RCED-96-180. 


      THE SPECIFIC SALES MECHANISM
      AND PROCESS NEED TO BE
      DETERMINED
-------------------------------------------------------- Chapter 3:3.5

The objectives underlying a divestiture help determine the most
appropriate sales method.  For example, if a divestiture is largely
motivated by fiscal considerations, an appropriate sales mechanism
would involve some form of competitive bidding and tend to place few
restrictions on the number or identity of bidders.\52 For example,
the Congress, in the 1996 National Defense Authorization Act,
directed DOE to sell its Naval Petroleum Reserve No.  1 (Elk Hills)
by February 1998 and to do so in a manner that would obtain the
maximum proceeds to the government.\53 The government has been
producing and selling oil and gas from the field for the past 20
years.  According to DOE, the reserve's sale is part of an effort to
remove the federal government from nonfederal functions.  In October
1997, DOE announced that it had executed agreements preparing for the
reserve's sale for $3.65 billion in cash as a result of a competition
designed to allow all qualified bidders to compete.  Before the final
selection, DOE had contacted more than 200 companies and received 22
bona fide offers, according to DOE.  This sale, which was finalized
on February 5, 1998, is the largest divestiture in U.S.  government
history, according to DOE.  In general, we have supported the
principle that the federal government should receive full market
value in selling its assets.\54 Alternatively, if the major
motivation of a divestiture is to transfer operations to the private
sector, the government could choose to negotiate a sales price with a
selected buyer. 

In practice, the size of the assets to be sold, in terms of value and
scale of enterprise, has influenced the type of sales process used. 
Trade sales and public stock offerings are general processes; trade
sales are used more often to sell smaller enterprises or assets and
public offerings to sell larger ones.  Sales can be organized using
competitive bidding methods or negotiations with either type of sale. 
A brief description of these processes follows: 

  -- "Trade sales" draw on the idea that an existing set of
     businesses competing in the relevant line of business (or trade)
     are likely to offer more and higher bids for the assets.  Three
     key attributes of the PMAs and the electricity industry may lend
     themselves to a trade sale:  (1) the PMAs and related hydropower
     assets are part of an established industry with capital market
     connections experienced in the valuation, grouping, and sale of
     electricity-generating assets; (2) sales of significant
     electricity-generating assets are not unusual; (3) several
     bidders are likely for at least large portions of the PMAs and
     their related assets, depending on how those assets are grouped
     for sale.  A trade sale can be a negotiated sales process
     between the government and a buyer or can be accomplished using
     an auction to determine both the sales price of the assets as
     well as buyers. 

  -- Stock offerings have been used domestically, most recently in
     the sale of Conrail in 1987, as well as internationally to
     divest large public enterprises.  This method of sale would most
     likely require creating a government corporation or corporations
     out of the PMAs and their associated assets.  Some of these
     assets could be grouped for sale, and some could be excluded
     from the sale, depending on the policy trade-offs discussed.  In
     the case of some federal water projects, for example, the
     government could decide to retain control of the dam and
     reservoir to satisfy increasingly significant restrictions on
     the use of water because of concerns about the environment or
     endangered species.  The stock of the government corporation
     would be subsequently sold through standard financial market
     methods, such as a private placement through negotiations
     between particular investors and the government or through a
     sale to the general public by using competitive bidding. 

In cases where auction methods may be used to sell government assets,
recent government experience indicates the importance of carefully
choosing the specific format for an auction.  That is, a policy
decision to choose a competitive auction format requires making many
subsequent decisions to define the specific rules leading to an
appropriate operational auction.  For example, the Federal
Communications Commission chose to auction the leases of
electromagnetic spectrum licenses for use in mobile communications. 
While generating a large amount of revenue was a less important
objective than achieving an efficient geographic allocation of
spectrum licenses to communications firms, the auctions generated
more revenue than some potential bidders had predicted, according to
auction analysts.  In large part, in structuring these auctions, the
government carefully considered the auction format and the
identification of particular problematic features of auctions of
similar assets in other nations. 

Most domestic and international divestitures have relied on private
capital market firms as consultants and managers because of their
frequent experience with complicated and high-valued transactions
governing the transfer of assets in the private sector.  Particularly
in the case of public offerings but also for trade sales, the
government would be likely to incur substantial costs to prepare its
assets for sale or to pay for services performed by its financial
advisers.  For example, in the sale of Conrail, the government
employed a variety of financial advisers and a prominent law firm
with expertise in a variety of fields, including tax and employment
law.  Also, legislation authorizing the sale of DOE's Elk Hills Naval
Petroleum Reserve required DOE to use an investment adviser to
administer the sale. 

If the government's objective is to perpetuate the social and public
policy compacts concerning public power, it could transfer or sell
its hydropower assets to the preference customers.  The assets could
be sold free of the debt associated with them.  Although such a
transaction would provide some revenue to the Treasury, it would
probably provide less of a return to the Treasury than a sale to
parties that would be willing to pay the highest bid possible for the
assets.  A debt-free transfer is also harmful to the Treasury because
it would incur the debt associated with the hydropower assets,
including perhaps any associated debt previously repaid by power
revenues--for example, the federal investment in irrigation projects
beyond the ability of irrigators to repay.  A variation of this
suboption is contained in a bill now before House committees. 
According to the bill's sponsor, this proposal is designed to avoid
the fight over elimination of preference by issuing warrants
entitling the existing preference customers to purchase, by a pre-set
date and at a stipulated price, a fixed number of shares (based on
recent electricity purchases) in the PMA from which they purchase
power.  The stipulated price would be set somewhat below the expected
market price value of the shares.  The warrants would be fully
negotiable so that the preference customers could sell them if they
so chose.  The actual sale of the shares would be made to
individuals, which could be IOUs or investment bankers, holding the
warrants on the specified day of sale. 


--------------------
\52 In general, because bids would be likely to increase with more
bidders, restrictions on the number of bidders would be likely to
lead to smaller sales proceeds.  A World Bank survey of international
experiences with divestitures indicates that open bidding among
competitors is preferable to sales that rely on negotiations with
selected bidders because competitive bidding offers less opportunity
for favored buyers to receive special treatment at the taxpayers'
expense. 

\53 An administration proposal to corporatize and sell the reserve in
fiscal year 1996 is discussed in Naval Petroleum Reserve: 
Opportunities Exist to Enhance Its Value to the Taxpayer
(GAO/T-RCED-95-136, Mar.  22, 1995). 

\54 See Lessons Learned About Evaluations of Federal Asset Sales
Proposals (GAO/T-RCED-89-70, Sept.  26, 1989). 


      IMPACT OF A DIVESTITURE ON
      PREFERENCE CUSTOMERS' RATES
      SHOULD BE CONSIDERED
-------------------------------------------------------- Chapter 3:3.6

How a divestiture could affect preference customers' rates needs to
be considered.  Some of Southeastern's, Southwestern's, and Western's
customers are concerned that a sale would significantly raise their
rates.  From 1990 through 1995, the three PMAs received less than 2
cents per kWh for their power--at least 40 percent less than what the
nonfederal utilities received per kWh during the same period. 
However, proponents of divestiture contend that competition in the
wholesale market would be likely to moderate rate increases.  For
example, representatives of the Edison Electric Institute (the trade
association for IOUs) maintain that because the wholesale market is
competitive, very few preference customers will lack access to
alternate power suppliers following a divestiture.  They believe
that, after a PMA is divested, some preference customers who relied
heavily on that PMA will be able to purchase power from independent
power producers, energy brokers, or energy marketers at competitive
rates.  In addition, as we noted earlier in this report, many states
are moving toward deregulating both wholesale and retail markets. 

Representatives of PMAs and their customers believe that having
access to alternate supplies of electricity is not enough.  They note
that even in cases in which preference customers may buy most of
their electricity from alternate sources, these customers often rely
on the PMA for power during hours of peak demand, particularly in
areas where Southeastern and Southwestern sell power.  Having access
to inexpensive power during times of peak demand is important to
these customers because, typically, power sold to meet this demand is
more expensive than power sold at other times.  In response, Edison
Electric Institute officials maintain that preference customers will
be able to purchase power even during peak periods at competitive
prices. 

To address these concerns, we estimated how much preference
customers' rates might increase if the PMAs were divested.  We
examined only the potential rate impacts of divesting the PMAs and
excluded other factors that are currently volatile and difficult to
project.  In our analysis, we assumed, among other things, that (1)
immediately after a divestiture, the buyer of the PMA would raise
each preference customer's rates to the level the customer paid for
non-PMA power in 1995 and (2) the preference customers do not change
the quantity of electricity they purchased in 1995.  Because of a
lack of data, we did not assess how increasing competition in the
wholesale market may affect the rate changes from divestiture.  Also,
we did not project whether the emergence of competition in retail
markets would affect rates in the wholesale market.  It is important
to note that our methodology yields conservative results.  If prices
for wholesale power decline in the future, as many industry analysts
believe they will, preference customers' actual rate changes from
divestiture will be smaller than our estimates. 

Our analysis shows that most preference customers will experience
relatively small rate increases after a divestiture of the PMAs.  As
shown in figure 3.1, we estimate that more than two-thirds of
preference customers may see rate increases of 25 percent or less, or
up to 0.5 cents per kWh.  If the preference customers passed these
costs directly on to their end-users, the average residential
end-users' electricity bills would increase by no more than $4.17 per
month.  However, we also estimate that some preference customers,
mainly those that purchase a large portion of their power from the
PMA, may see their rates increase more.  About 13 percent of
preference customers may see rate increases that exceed 75 percent. 
Expressed in kWh, about 16 percent of preference customers may see
their rates increase by more than 1.5 cents per kWh.  If costs are
passed directly, the average residential end-users served by about 25
percent of preference customers would see their electricity bills
increase by more than $8.33 per month. 

   Figure 3.1:  Projected Rate
   Changes After a Divestiture for
   the Preference Customers of
   Southeastern, Southwestern, and
   Western

   (See figure in printed
   edition.)

Source:  GAO's analysis of data provided by EIA, Southeastern,
Southwestern, and Western. 

Preference customers who currently purchase a small portion of their
total power from Southeastern, Southwestern, or Western generally may
experience smaller rate increases after a divestiture.  For example,
in fiscal year 1995, 99 percent of Southeastern's preference
customers received less than a quarter of their power from the PMA. 
Correspondingly, as illustrated in figure 3.2, we calculated that
almost all (98 percent) of Southeastern's preference customers may
experience rate increases of 0.5 cents per kWh or less, and 99
percent would see their rates increase by one-quarter or less. 
Moreover, we estimated that about 27 percent (or 72) of these
customers may see their rates decline if they purchased all of their
power at 1995 wholesale market rates.  Some of these customers
currently may have access to less expensive power; however, for
various reasons, these customers have opted not to buy from these
sources.\55

   Figure 3.2:  Post-Divestiture
   Rate Changes for Southeastern's
   Preference Customers

   (See figure in printed
   edition.)

Source:  GAO's analysis of data provided by EIA and Southeastern. 

In contrast, preference customers who currently purchase most or all
of their power from the PMA may experience much greater rate
increases.  For example, in 1995, about 38 percent of Western's
preference customers purchased more than half of their electricity
from the PMA.  As shown in figure 3.3, we estimated that about
one-fifth of Western's customers may see their rates increase by more
than 75 percent.  About 27 percent of preference customers may see
rate increases greater than 1.5 cents per kWh.  If preference
customers pass the higher rates on to those they serve, the average
residential end-users served by about 16 percent of Western's
preference customers may see their electricity bills increase by at
least $16.67 per month. 

Similarly, almost one-third of Southwestern's preference customers
purchase more than 75 percent of their electricity from the PMA.  As
shown in figure 3.4, although most of Southwestern's preference
customers will experience relatively small rate changes, about 25
percent may see their rates more than double.  If these preference
customers pass these increases on to those they serve, the average
residential end-users may see their rates increase by at least $16.67
per month. 

   Figure 3.3:  Post-Divestiture
   Rate Changes for Western's
   Preference Customers

   (See figure in printed
   edition.)

Source:  GAO's analysis of data provided by EIA and Western. 

   Figure 3.4:  Post-Divestiture
   Rate Changes for Southwestern's
   Preference Customers

   (See figure in printed
   edition.)

Source:  GAO's analysis of data provided by EIA and Southwestern. 

It is important to remember that, although some preference customers
may initially experience significant rate increases, government may
mitigate these rate increases through various mechanisms, such as
rate caps.  In addition, these customers currently pay rates that, on
average, are 40 to 50 percent below what neighboring utilities pay
that do not have access to PMA power.  After the divestiture, these
preference customers will be paying the same market rates as those
utilities. 

Finally, smaller-sized preference customers may experience larger
rate increases after divestiture.\56 As illustrated in figure 3.5, we
estimated that about one-fifth of Southeastern's, Southwestern's, and
Western's small preference customers will experience rate increases
exceeding 75 percent.  About 30 percent of small customers will see
their rates rise by more than 1.5 cents per kWh.  In contrast, 2
percent of medium-sized preference customers and 3 percent of large
preference customers may see rate increases exceeding 75 percent. 
However, in all three size categories, a majority of preference
customers may experience rate increases of 25 percent or less or 0.5
cents per kWh or less.  We believe smaller customers may experience
larger rate increases after divestiture because they generally
purchase a larger portion of their power from the PMAs than
medium-sized and large preference customers. 

   Figure 3.5:  Projected Rate
   Changes After a Divestiture for
   Southeastern's, Southwestern's,
   and Western's Preference
   Customers, by Size of Customer

   (See figure in printed
   edition.)

Source:  GAO's analysis of data provided by EIA, Southeastern,
Southwestern, and Western. 


--------------------
\55 The customers that may experience a rate decrease are those that
are currently purchasing power from the PMA at rates that are above
the market price.  In theory, in these situations, after a
divestiture, the rate for power formerly provided by the PMA would
decrease to the prevailing market rate, and these customers would
experience an overall decrease in the cost of their power.  However,
according to PMA customers, this analysis does not consider the fact
that the PMA's power, in many cases, satisfies demand during peak
periods.  According to PMA customers, in this niche, the PMAs' power
is often less expensive than peaking power offered by other sources. 
Some PMA customers have built their own generating capacity based on
buying a PMA's power and using it for peaking purposes.  They
maintain that it would be costly and difficult to replace the power
supplied by the PMAs because it is unlikely that less expensive
sources of power could be found for peaking purposes and that they
may be forced to build new types of baseload capacity if their
resource requirements change.  They also do not believe that a buyer
of the PMA would necessarily decrease the price of the PMA's power to
match overall power rates but would be more likely to increase the
price to match that of power generated from power plants used to
serve peak demand. 

\56 We measured size by the number of MWh that each preference
customer delivered to its end-users from all sources in calendar year
1995.  We categorized size as follows:  "small" = 0 to 100,000 MWh;
"medium" = more than 100,000 to 500,000 MWh; "large" = more than
500,000 MWh.  We discussed these categories with the American Public
Power Association, an association of publicly owned utilities, and
the National Rural Electric Cooperatives Association, an association
of consumer-owned rural electrical systems. 


RESULTS OF GAO'S PRIOR WORK ON THE
TENNESSEE VALLEY AUTHORITY
=========================================================== Appendix I

The Tennessee Valley Authority (TVA) had $27.9 billion of debt and
$6.3 billion of deferred assets on September 30, 1996.  In reports we
issued in 1995\1 and 1997,\2 we concluded that TVA's high fixed costs
and deferred assets may hinder its ability to compete if TVA is
required to participate in a deregulated market.  In a competitive
market, where wholesale prices are expected to decrease, TVA's high
fixed costs and deferred assets make it reasonably possible that the
federal government would incur future
losses.\3 However, in recent years, TVA, the nation's largest
electric power generator, has taken several actions to improve its
competitiveness.  In addition to reducing its labor force,
refinancing its debt, and bringing two deferred nuclear units back
into service, TVA has recently increased its rates as part of its
efforts to reduce its debt by 50 percent by fiscal year 2007.  TVA's
service area is protected from competition under federal law; as long
as this is the case, the risk that TVA will cause the federal
government to incur losses is remote. 


--------------------
\1 Tennessee Valley Authority:  Financial Problems Raise Questions
About Long-term Viability (GAO/AIMD/RCED-95-134, Aug.  17, 1995). 

\2 Federal Electricity Activities:  The Federal Government's Net Cost
and Potential for Future Losses:  Volume 1 (GAO/AIMD-97-110, Sept. 
19, 1997). 

\3 We based our discussion of the risk of nonrecovery involved on the
Statement of Federal Financial Accounting Standard No.  5, Accounting
for Liabilities of the Federal Government.  The Standard states that
if the chance a contingent loss will occur is more likely than not,
the risk of loss is "probable"; if the chance is more than remote but
less than probable, it is "reasonably possible"; if the chance is
slight, it is "remote."


   BACKGROUND
--------------------------------------------------------- Appendix I:1

TVA was established by the Tennessee Valley Authority Act of 1933 as
a multipurpose, independent, federal corporation.  The act created
TVA to improve the quality of life in the Tennessee River Valley by
improving navigation, promoting regional agricultural and economic
development, and controlling the flood waters of the Tennessee River. 
As part of TVA's efforts to fulfill these objectives, it erected dams
and hydropower facilities on the Tennessee River and its tributaries. 
TVA also developed fertilizers, taught farmers how to improve crop
yields, and helped replant forests, control forest fires, and improve
habitats for wildlife and fish. 

To meet the growing need for electric power during World War II, TVA
quickly expanded its construction of hydropower plants.  By the end
of the war, TVA had become the nation's largest electricity supplier. 
However, the demand for electricity in the region outpaced TVA's
capacity.  To secure funding for the construction of coal-fired power
plants, TVA sought the authority to issue bonds.  The Congress passed
legislation in 1959 that gave TVA the authority to issue bonds and
required TVA's power program to be self-financed.\4 TVA's debt limit
is set by the Congress and was established at $750 million in 1959.\5

The 1960s was a period of unprecedented economic growth in the
Tennessee Valley.  Expecting the Valley's electric power needs to
continue to grow, TVA decided to add nuclear power plants to its
power system.  In 1996, TVA had a dependable generating capacity of
over 28,000 megawatts (MW).  The system primarily consists of 113
hydroelectric units, 59 coal-fired units, 48 combustion turbines, and
5 operating nuclear units. 

TVA's power program generated $5.7 billion in revenues in fiscal year
1996.  As of January 1998, TVA sells power at wholesale rates to 159
municipal and cooperative distributors and to a number of directly
served large industrial customers and federal agencies.  TVA's sales
to its distributors in fiscal year 1996 constituted approximately
$5.0 billion (or 88 percent) of TVA's total revenue for the year. 
Most of the power contracts between TVA and its distributors contain
a 20-year term that automatically renews each year and require that
the distributors give TVA at least a 10-year notice of cancellation. 
The distributors, in turn, sell the power to nearly 8 million people
in an 80,000-square-mile area covering Tennessee and parts of
Alabama, Georgia, Kentucky, Mississippi, North Carolina, and
Virginia. 


--------------------
\4 TVA's activities are divided into two types--the power program and
the nonpower programs.  The nonpower programs, such as flood control,
navigation, and water resources, are primarily funded through federal
appropriations and user fees.  The nonpower programs received $106
million in funding for fiscal year 1997 and operate primarily within
the 41,000-square-mile Tennessee River watershed. 

\5 Since then, the Congress has raised the debt limit to $30 billion. 


   TVA'S FINANCIAL CONDITION
   REDUCES ITS FLEXIBILITY AND
   ABILITY TO COMPETE IN THE
   FUTURE
--------------------------------------------------------- Appendix I:2

As we discussed in reports issued in 1995 and 1997,\6 TVA's high
debt, related financing costs, and deferred assets would limit the
agency's flexibility to respond to competitive pressures if it were
no longer protected from competition.  TVA has operated with little
oversight in the past, and investments in its construction program
for nuclear power plants constitute most of its debt and all of its
deferred assets. 


--------------------
\6 GAO/AIMD/RCED-95-134 and GAO/AIMD-97-110. 


      DESIGNED AS A WHOLLY OWNED
      GOVERNMENT CORPORATION, TVA
      OPERATES WITH LITTLE OUTSIDE
      OVERSIGHT
------------------------------------------------------- Appendix I:2.1

TVA's authorizing legislation allows it to operate with a high degree
of independence.  The TVA Act of 1933 did not subject TVA to the
regulatory and oversight requirements that must be satisfied by
commercial electric utilities.  For example, unlike other utilities,
TVA's power rates and power resource decisions are not subject to
review and approval by state public utility commissions or the
Federal Energy Regulatory Commission (FERC). 

Instead, all authority over TVA's operations is vested in TVA's
three-member board of directors, including the sole authority to set
wholesale power rates and approve the retail rates charged by TVA's
distributors.  The three board members are full-time TVA employees. 
They are appointed by the President, with the advice and consent of
the Senate, and serve 9-year, overlapping terms of office.  The
President designates one member as the chairman.  In addition, the
Congress has little oversight over the funding of TVA's power
program, which is self-financed through power revenues and bond
issuances and does not require federal appropriations.  TVA's power
funds are maintained in a revolving fund called the TVA Fund. 

The issue of TVA's oversight has been examined several times in the
past.  For example, in a 1982 report, we pointed to a growing concern
with TVA's activities and identified options for improving oversight
and accountability.\7 These options included periodic congressional
oversight hearings and placing the TVA rate-setting process under
FERC.  In a 1983 report, we discussed our concerns about TVA's
management and concluded that the issue of the adequacy of TVA's
oversight needed greater attention.\8 In a 1987 report entitled
"TVA--A Path to Recovery," the Southern States Energy Board concluded
that "additional mechanisms are needed to ensure that TVA is
accountable for its actions to its ratepayers, Congress, and the
American public."\9 The report further stated that: 

     "There must be a fundamental change in TVA's structure to
     effectively respond to today's challenges and meet the necessary
     standards of accountability.  A larger Board should be
     established, comprised of part-time directors who would be
     responsible for policy-making and oversight of TVA's
     management."

In 1997, TVA's oversight was a topic of debate in the Congress.  The
possibility of deregulating electric utilities in the future led one
Representative to propose the formation of an independent regional
commission to make recommendations to the President and the Congress
on a strategy for TVA's future in a deregulated environment.  Another
Congressman has expressed interest in the expansion of TVA's current
board of directors.  In October 1997, a bill was introduced in the
Senate to expand TVA's board from three full-time members to nine
part-time members, each having a strong background in corporate
management or strategic decision-making.  Under this proposal, the
expanded board would establish long-range goals and policies for TVA,
but it would leave the day-to-day management to an independent chief
executive officer.  According to the bill's sponsor, such a
management structure could help TVA "avoid the type of decisions and
missteps that have saddled TVA with more than $27 billion in debt
over the years" and "can help this important agency face the upcoming
dramatic changes in the electric utilities industry as effectively
and efficiently as possible." As of November 30, 1997, neither
proposal had been implemented. 


--------------------
\7 Tennessee Valley Authority--Options for Oversight (GAO/EMD-82-54,
Mar.  19, 1982). 

\8 Triennial Assessment of the Tennessee Valley Authority--Fiscal
Years 1980-1982 (GAO/RCED-83-123, Apr.  15, 1983). 

\9 The Southern States Energy Board was comprised of government and
industry experts with diverse experience in energy operations,
management, and regulation. 


      INVESTMENT IN NUCLEAR
      PROGRAM INCREASED DEBT
------------------------------------------------------- Appendix I:2.2

TVA made its commitment to nuclear power in the late 1960s and early
1970s, when power sales were growing at a steady rate and were
expected to double every 10 years.  By 1970, TVA customers used
nearly twice as much electricity as the national average.  At that
time, TVA was experiencing an annual growth rate of about 8 percent
in demand for electricity, and its forecasts through the mid-1970s
were showing continued high growth in demand. 

In 1966, TVA announced plans to construct a total of 17 nuclear units
at seven sites in Alabama, Mississippi, and Tennessee to satisfy its
forecast demand.  However, instead of increasing, electricity
consumption declined in the mid-1970s following the 1973 energy
crisis and again in the late 1970s with higher energy costs and lower
economic growth.  In addition, because of the Three Mile Island
nuclear accident in 1979, the Nuclear Regulatory Commission (NRC)
issued extensive new safety regulations that applied to all nuclear
plants.  The decreasing demand for electricity, coupled with the
increased regulation of nuclear power, caused the electric utility
industry to rethink the role that nuclear power would play in meeting
the nation's demand for electricity.  Most utilities chose to cancel
ongoing nuclear construction projects as well as planned nuclear
power plants. 

After reassessing its electricity demand forecasts using a more
sophisticated methodology, TVA began to scale back its nuclear plans
by canceling 8 of its 17 planned nuclear units in 1982 and 1984.  The
almost $5 billion invested in these eight units was written off over
10 years and recovered through rates.  TVA's remaining nine nuclear
units have had a long history of operating and construction problems. 
As of September 30, 1996, TVA had five nuclear units in operation. 
The two most recent additions to TVA's nuclear power resources are
Browns Ferry 3, which was returned to service in November 1995, and
Watts Bar 1, which began commercial operation in May 1996.  Browns
Ferry 3 began operations in 1977 but was shut down in 1985 because of
repeated operational and maintenance errors.  Watts Bar 1 had been
under construction for about 23 years and had never been operated. 
Construction at these two nuclear units involved years of schedule
delays and cost overruns.  For example, TVA certified to the NRC that
Watts Bar 1 was qualified for an operating license in 1985, but the
Commission did not grant one because of over 5,000 unresolved
concerns about construction deficiencies and management practices at
the facility that were reported by TVA employees.  According to TVA,
the total costs associated with the completion of Watts Bar 1 and
Browns Ferry 3 were about $6.9 billion and $1.4 billion,
respectively, as of September 30, 1996. 

Of TVA's four remaining nuclear units, Bellefonte 1 and 2 and Watts
Bar 2, were not completed and have been kept in a "mothballed"
status.  In December 1994, TVA determined that it would not, by
itself, complete Bellefonte 1 and 2 or Watts Bar 2 as nuclear units. 
TVA has been considering the possible conversion of the Bellefonte
plant to a combined cycle plant utilizing another fuel source, such
as gas or coal, and/or the formation of a joint venture with a
partner for completion of the plant.  TVA also concluded that Watts
Bar 2 should remain in deferred status until TVA completes the
Bellefonte study.  TVA has already invested about $6.3 billion in
these three units.  The remaining "mothballed" nuclear unit, Browns
Ferry 1, has been shut down since 1985 because of ineffective
management and technical difficulties.  While TVA's investment in
Browns Ferry 1 totals approximately $86 million, it "will continue to
remain in an inoperative status until its ultimate disposition is
determined," according to TVA's fiscal year 1996 annual report. 

Despite the past problems TVA experienced with its nuclear program,
TVA has recently reported positive developments concerning its
nuclear units.  In 1996, the NRC conducted performance reviews of
Watts Bar 1 and the two operating units at Browns Ferry.  The NRC
gave either "good" or "superior" rankings to the three units in the
four functional areas of engineering, maintenance, operations, and
plant support.  Nevertheless, while five of TVA's nine nuclear units
are operational, TVA's investment in its nuclear power program has
left it in a difficult financial condition that may limit its ability
to compete in a deregulated market. 


      TVA'S FINANCIAL CONDITION
      MAY LIMIT ITS FLEXIBILITY
------------------------------------------------------- Appendix I:2.3

Primarily as a result of TVA's investment in nuclear power, TVA's
outstanding debt grew from $15 billion at the end of fiscal year 1983
to almost $28 billion at the end of fiscal year 1996.  The
outstanding debt consists primarily of about $3.2 billion in direct
federal borrowing from the Federal Financing Bank and about $24.1
billion in publicly issued TVA debt, which is not explicitly
guaranteed by the federal government.  In addition, TVA is also
required to repay funds appropriated to it prior to its becoming
self-funding in 1959--approximately $600 million as of September 30,
1996.\10

As a result of its debt, TVA's total interest expense in fiscal year
1996 was about $2 billion, representing about 35 percent of TVA's
operating revenue, according to TVA's annual report.  TVA's ratio of
financing costs to revenue is now more than twice as high as the
average financing costs for neighboring utilities.  In addition,
TVA's ratio of fixed financing costs to revenue is almost five times
higher than the average of its neighboring investor-owned utilities
(IOUs).  The high debt and high financing costs allow TVA less
flexibility to reduce costs and, hence, to lower its rates to meet
competitors' prices. 

In addition, in September 1997 we reported that TVA deferred about
$6.3 billion in capital costs for its nonproducing nuclear assets to
future years rather than currently including them among the costs
being recovered from ratepayers.\11 TVA considers these assets--the
Bellefonte 1 and 2 and Watts Bar 2 nuclear units--to be construction
work-in-progress.  TVA has concluded that the recovery of the costs
of these assets will not begin until the units are either completed
and placed into service or canceled.  TVA charges its ratepayers for
the costs of its property, plant, and equipment and canceled plants
through depreciation and amortization expenses.  TVA is required by
law to set rates so that power revenues cover all operating expenses,
including depreciation and amortization.  While the annual interest
expense from the debt associated with these assets is included in
current rates, TVA is not currently depreciating or amortizing its
nonproducing nuclear assets.  TVA has stated that it will not, by
itself, complete Bellefonte 1 and 2 or Watts Bar 2 as nuclear units,
and it has not conducted any construction work on these units for
approximately 9 years.  As we reported previously, we believe that
the $6.3 billion in costs associated with these three units does not
represent viable construction projects.\12 These are the only
deferred nuclear units in the United States.  In our judgment, it is
no longer reasonable for these costs to be deferred from current
revenue requirements.  How much TVA's revenue requirements will
increase depends on when and over what period of time TVA begins
recovering its investment in its nonproducing nuclear assets.  By not
including the costs of its deferred nuclear units in rates and using
the cash to pay off debt in prior years, TVA has allowed its high
fixed and deferred costs to put upward pressure on its rates at a
time when competitors' power rates are expected to be falling. 


--------------------
\10 TVA refers to this debt as "appropriated investment"; however,
this amount does not count toward TVA's $30 billion debt limit.  TVA
must repay all but $258.3 million of the appropriations that were
used for capital investments, plus interest. 

\11 GAO/AIMD-97-110. 

\12 GAO/AIMD-97-110. 


      ELECTRICITY INDUSTRY IS
      BECOMING MORE COMPETITIVE
------------------------------------------------------- Appendix I:2.4

As previously mentioned, IOUs have historically maintained exclusive
service areas in return for providing electric service to all
customers in their areas.  Through their electricity rates, the IOUs
generally recoup the costs to build new generating plants and to
operate the power system plus a regulated return.  In 1959, the
Congress legislatively defined TVA's service territory.  However,
recent changes in the electricity industry are pushing utilities
closer to a competitive market, and utilities have been forced to
adopt a more competitive strategy to survive. 

The Energy Policy Act of 1992 promoted increased competition in the
electricity market.  The act encouraged open transmission of
electricity by allowing wholesale electricity customers, such as
municipal distributors, to purchase electricity from any supplier,
even if that power had to be transmitted over lines owned by another
utility.  In addition, bills have been introduced in various House
and Senate committees to promote or mandate retail electricity
competition, and several states are actively implementing retail
competition.  State regulators hope that industrial, commercial, and,
ultimately, residential consumers will be able to choose their power
supplier. 

In the light of the recent push toward deregulation and competition,
utilities have begun to adopt new strategies to compete.  Some are
acquiring or merging with other utilities in order to better respond
to market changes.  Others are investing in different industries,
such as home security and telecommunications.  Utilities are also
restructuring themselves and decreasing their operating costs through
reorganizations and layoffs.  Utilities have implemented these and
other strategies in response to the uncertainties about the future of
the electricity markets. 

While the Energy Policy Act exempted TVA from the act's
transmission-related requirements, thus preventing competitors from
using TVA's transmission system to sell power to customers inside
TVA's service area, some of TVA's customers have recently expressed
interest in buying power from other sources, to the point of wanting
to leave TVA's power system altogether.  For example: 

  -- In December 1993, the Four-County Electric Power Association of
     Columbus, Mississippi, announced that it was canceling its
     contract with TVA, effective in December 2003.  Four-County
     officials said that a study they commissioned indicated that
     TVA's wholesale rates may increase by 30 percent over a 10-year
     period.  By buying power from sources other than TVA,
     Four-County believed it could reduce its power costs by about 25
     percent.  TVA threatened to cancel plans to construct a
     lignite-burning power plant in Four-County's region if it did
     not withdraw its cancellation notice.  In May 1996, Four-County
     withdrew the notice and agreed not to give a 10-year
     cancellation notice for the next 5 years. 

  -- In Virginia, the Bristol Utilities Board left the TVA system for
     Cinergy Corp., effective January 1, 1998.  Cinergy offered
     Bristol firm wholesale power at 2.59 cents per kWh for 7
     years--40 percent less than TVA's wholesale rate of 4.3 cents
     per kWh.  According to its general manager, Bristol would save
     $70 million over 7 years.  Bristol, which is on the border of
     TVA's service area, has the ability to pursue the agreement with
     Cinergy because it does not have a long-term contract with TVA. 
     Bristol also received a unique exemption in the Energy Policy
     Act of 1992 that allows other utilities to transmit electricity
     to Bristol over TVA's power lines.  While Cinergy may have
     offered this power to Bristol at marginal rates, this is the
     type of competitive situation that TVA may face regularly if it
     loses its current protection from competition.  TVA is
     attempting to recover stranded investment costs from Bristol. 

  -- In May 1997, the board members of the Paducah Power System in
     Kentucky voted to give TVA a 10-year notice of intent to cancel
     Paducah's contract with TVA.  The board had been presented with
     at least one study showing that Paducah could buy power from
     other sources for 10 to 15 percent less than the amount that
     they were paying TVA.  The proposed change must be approved by
     the Paducah City Council. 

  -- The five largest distributors in TVA's system--Huntsville,
     Chattanooga, Knoxville, Nashville, and Memphis--have expressed
     concern about the inflexibility of TVA's power contracts.  These
     utilities, which account for more than one-third of TVA's
     distributor power sales, hired a consultant to help develop
     proposals to present to TVA.  These distributors are interested
     in contract flexibility through the negotiation of shorter
     contracts and favor the ability to purchase power from outside
     sources.  These large distributors anticipate using their
     leverage to compel TVA to renegotiate their power contracts. 


   FEDERAL GOVERNMENT FACES
   DIFFERENT DEGREES OF RISK WITH
   TVA
--------------------------------------------------------- Appendix I:3

Despite the industry's push toward competition and pressure from some
of TVA's customers, several factors protect TVA from competition,
making the risk of loss to the federal government remote in the short
term.  The long-term risk, however, appears to be greater. 

The federal government's financial exposure from TVA is nearly $28
billion because of its direct and indirect financial involvement. 
The risk that TVA will cause the federal government to incur losses
is remote as long as TVA retains a position in its service area that
is protected from competition.  However, if TVA loses its protected
position and is required to compete at a time when wholesale prices
are expected to be falling, its high financing costs and deferred
assets make it reasonably possible that the federal government could
incur losses in the future.  The federal government's direct
financial involvement with TVA consisted of about $600 million of
appropriated debt\13 and about $3.2 billion in Federal Financing Bank
debt, as of September 30, 1996.  If TVA fails to make future payments
on its outstanding appropriated and Federal Financing Bank debt, the
federal government will incur a loss.  The government could also
incur a loss because of its indirect financial involvement, which
consists of TVA's public debt of about $24.1 billion, as of September
30, 1996, should it have to absorb unreimbursed costs from any
actions it would take to prevent default on the debt service
requirements. 


--------------------
\13 TVA's appropriated debt consists of appropriations that were
primarily used to construct TVA's hydroelectric and fossil plants,
transmission system, and other general assets of the power program. 
TVA must make annual principal payments (currently $20 million) to
the Treasury from net power proceeds plus a market rate of return on
the balance of this debt. 


      TVA'S PROTECTION FROM
      COMPETITION MAKES FEDERAL
      GOVERNMENT'S SHORT-TERM RISK
      REMOTE
------------------------------------------------------- Appendix I:3.1

Two major factors protect TVA from competition and allow it to
operate in a manner similar to a traditionally regulated electric
utility monopoly.  First, in nearly all instances, TVA's contracts
with its 159 distributors require the distributors to give at least a
10-year notice before they can switch to another power supplier.\14
Second, TVA is exempt from the transmission-related provisions of the
Energy Policy Act of 1992.  This exemption prevents other utilities
from using TVA's transmission system to sell power to customers
inside TVA's service area. 

TVA's wholesale contracts with its distributors are generally
long-term contracts that ensure TVA a relatively stable customer base
and cash flow.  These contracts represented about 83 percent of TVA's
load as of September 30, 1996.  Most of the wholesale power contracts
between TVA and its distributors contain a 20-year term that
automatically renews each year (referred to as the "evergreen"
provision) and require that the distributors give TVA at least a
10-year notice of cancellation.  This notice provision effectively
locks the distributors into purchasing power from TVA since obtaining
price quotes for power to be supplied 10 to 15 years into the future
is generally not feasible.  All of the power contracts between TVA
and its distributors are "full requirements" contracts, which require
the distributors to purchase all of their electric power from TVA. 

TVA is further insulated from competition by a specific exemption
from the transmission-related provisions of the Energy Policy Act of
1992.  Under the act, FERC can compel a utility to transmit
electricity generated by another utility into its service area for
sale to wholesale customers.  The act acknowledged that TVA is
legally prohibited from selling power outside its legislatively
mandated service area and therefore exempts TVA from having to
transmit power from neighboring utilities to wholesale customers
within TVA's service area.  While TVA is authorized to allow other
utilities to use its transmission lines to transmit power through its
service area to other utilities, it is not required to allow other
utilities to sell power to customers within TVA's service area. 


--------------------
\14 Some wholesale power contracts between TVA and wholesale
customers require a 15-year notice of cancellation. 


      RISK OF LOSS IS REASONABLY
      POSSIBLE IF TVA'S PROTECTION
      FROM COMPETITION ENDS
------------------------------------------------------- Appendix I:3.2

According to our discussions with industry experts and TVA officials,
it appears unlikely that TVA will be allowed to maintain its current
regulated monopoly-type structure indefinitely:  At some future
point, TVA will have to compete with other utilities.  In a
competitive environment, utilities that have low costs and the
flexibility to adjust their rates to meet those being offered by
other utilities are expected to be the most successful.  We believe
TVA's substantial fixed costs and deferred assets will limit TVA's
flexibility to continue to offer competitive rates and may affect its
ability to recover all costs when competitors' prices are being
driven down. 

TVA has chosen to defer costs from its substantial nuclear investment
to future years rather than include them in the current costs being
recovered from ratepayers.  As a result, TVA had accumulated about
$28 billion of debt, as of September 30, 1996, which resulted in
almost $2 billion in interest expense in fiscal year 1996.  The
recovery of these deferred assets is most likely to be scheduled at a
time when wholesale power rates are expected to be falling. 

In a previous report, we compared the financial ratios of TVA and
neighboring IOUs that indicate the flexibility of these entities.\15
We also computed ratios that compare the magnitude of TVA's deferral
of costs with those of its most likely competitors.  We found that
TVA's ratios of financing costs to revenue greatly exceed the ratios
of its neighboring utilities, indicating that TVA has less
flexibility to lower prices to meet competition.  In addition, the
calculation of deferred asset ratios indicated that while TVA has
deferred substantial costs, its potential competitors have written
down the assets they deem to be uneconomical at a much faster rate,
allowing them to recover costs at a much greater pace than TVA and
thus giving them greater financial flexibility in the future. 

The primary component of TVA's deferred assets is about $6.3 billion
in capital costs for its nonproducing nuclear assets--Bellefonte 1
and 2 and Watts Bar 2.  In December 1994, TVA determined it would
not, by itself, complete Bellefonte 1 and 2 or Watts Bar 2 as nuclear
units.  However, TVA is studying the potential for converting the
Bellefonte facility to a combined cycle plant or forming a joint
venture with a partner for completion of the plant.  This study was
scheduled to be completed by the fall 1997.  TVA also concluded, as
part of its Integrated Resource Plan, that Watts Bar 2 should remain
in deferred status until completion of the Bellefonte study. 

We believe that two additional factors could contribute to TVA's
future vulnerability to competition:  the concentration of TVA's
sales to its five largest distributors and the number of TVA's
customers that are already connected to the transmission lines of
other utilities.  As previously reported, the five biggest
distributors in TVA's system, which accounted for 34 percent of TVA's
total sales to distributors in fiscal year 1996, have expressed
concerns about their lack of flexibility to purchase power from
outside sources.  The large distributors hope to use their leverage
in order to compel TVA to renegotiate their power contracts.  In a
competitive environment, TVA would likely have to lower the rates of
these distributors or run the risk of losing them as customers, which
could be financially crippling to TVA.  In addition, 12 other TVA
distributors are already interconnected with other utilities.  These
distributors could get power from other sources after their contracts
with TVA expire.  The demand from these customers amounts to about 2
percent of TVA's total load.  As competition intensifies in the
region, TVA could lose distributors to other suppliers using existing
and future transmission connections. 


--------------------
\15 GAO/AIMD/RCED-95-134. 


   MITIGATING FACTORS REDUCE THE
   RISK OF LOSS
--------------------------------------------------------- Appendix I:4

Other factors, such as the inherent cost advantages of a federal
corporation and an extensive transmission system, mitigate the risk
created by TVA's high financing costs and deferred assets.  In
addition, TVA's management has taken several actions in recent years
to reduce TVA's expenses and make it more competitive.  Because of
these factors and actions, we believe the risk of loss to the federal
government is reduced but is still reasonably possible. 


      TVA HAS INHERENT COST
      ADVANTAGES
------------------------------------------------------- Appendix I:4.1

According to bond-rating agencies, TVA's creditworthiness is based on
its links to the federal government.  In accordance with the TVA Act,
TVA's debt issuances explicitly state on the bond prospectus that the
bonds are neither legal obligations of, nor guaranteed by, the
federal government.  Nevertheless, TVA's bonds are rated by the major
bond-rating agencies as if they have a federal guarantee.  Without
the links to the federal government, we believe that TVA would have a
lower bond rating and higher cost of funds. 

In addition, as a federal government corporation, TVA is exempt from
federal and state income taxes and does not pay various local taxes. 
While TVA is required to make payments in lieu of taxes to state and
local governments of the jurisdictions where power operations are
conducted, the base amount TVA is required to pay amounts to only
about 5 percent of TVA's gross power revenues (not including sales to
other federal agencies).\16 In addition, according to TVA, its
distributors are required to pay various state and local taxes, which
amounted to about $125 million, or about 2 percent of the total
fiscal year 1995 operating revenues of TVA and the distributors.  In
comparison, IOUs pay about 14 percent of their operating revenues for
taxes.  In addition, interest income for TVA's bondholders is
generally exempt from state income taxes, which further lowers TVA's
costs of funds. 

Other cost advantages that TVA possesses are its hydropower assets
and its preference in purchasing low-cost power from the Southeastern
Power Administration (Southeastern).  TVA has relatively more
hydroelectric power than neighboring utilities.  About 11 percent of
its power is generated from its 113 hydroelectric units at 29
conventional dams.  In comparison, an average of 6 percent of the
power from other utilities comes from hydroelectric dams.  These
established hydroelectric projects are relatively inexpensive and
have no associated fuel costs.  TVA also purchases about 2 percent of
its annual power needs from Southeastern.  In fiscal year 1996, TVA
purchased this power for 0.8 cents per kWh. 


--------------------
\16 In fiscal year 1996, for example, TVA made $256 million in
payments in lieu of taxes to state and local governments. 


      TVA'S RECENT ACTIONS HAVE
      LOWERED COSTS AND INCREASED
      REVENUES
------------------------------------------------------- Appendix I:4.2

Over the years, TVA has taken several steps to enhance its
competitiveness.  For example, it canceled a number of its nuclear
construction projects in the early 1980s and, more recently,
completed the construction of Watts Bar 1 and restarted Browns Ferry
3.  TVA also recently announced that it has internally capped its
debt limit at about $28 billion and plans to finance its future
capital expenditures from operations.  In addition, by reducing its
workforce from 34,000 in 1988 to 15,308 in June 1997 and refinancing
its debt at lower interest rates, TVA has reduced its annual
operating costs. 

In July 1997, TVA released a 10-year business plan that identifies
actions it plans to take to meet the challenges from the
restructuring electricity marketplace.  The proposed actions address
several of the concerns that we raised in our August 1995 report. 
The plan calls for TVA to

  -- increase power rates enough to increase annual revenues by about
     5.5 percent ($325 million);

  -- take various actions to reduce its total cost of power by about
     16 percent by fiscal year 2007;

  -- reduce employment levels to 14,275 by September 30, 1997;

  -- limit annual capital expenditures to $595 million; and

  -- reduce debt by about 50 percent from $27.9 billion, as of
     September 30, 1996, to $13.8 billion by fiscal year 2007. 

To the extent that TVA is able to use the cash generated from
increasing rates, reducing expenses, and capping future capital
expenditures to pay down debt, the risk of loss to the federal
government is reduced.  In addition to these actions, the plan calls
for TVA to change the length of the wholesale power contracts with
its distributors from a rolling 10-year term to a rolling 5-year term
beginning 5 years after the amendment.  However, reducing the length
of the wholesale contracts with its distributors could increase the
risk of loss to the federal government by giving TVA's customers more
flexibility to end their contracts with TVA. 


RESULTS OF GAO'S PRIOR WORK ON THE
RURAL UTILITIES SERVICE
========================================================== Appendix II

In September 1997,\1 GAO found that the Rural Utilities Service (RUS)
operates its loan programs at a net cost to the federal government
because the annual interest income received from RUS borrowers is
substantially less than the government's annual interest expense to
provide the funds to borrowers.  In addition, in fiscal years 1996
and 1997, RUS wrote off $1.6 billion in electric loans.  Moreover, as
of September 30, 1996, $10.5 billion of the $32.3 billion total
electric portfolio represented loans to borrowers that are bankrupt
or otherwise financially stressed.  As the electric utility industry
moves toward deregulation, it is probable that the federal government
will continue to incur substantial losses from financially stressed
borrowers and from other borrowers with high production costs and the
inability to raise rates because of regulatory and/or market
pressures. 

RUS, an agency within the Department of Agriculture, provides direct
and guaranteed loans primarily to rural electric cooperatives that
market power on a wholesale and retail basis.  Through RUS, the
Department of Agriculture, as the federal government's principal
provider of loans to assist the nation's rural areas in developing
their utility infrastructure, finances the construction, improvement,
and repair of electrical systems.  RUS provides credit assistance
through direct loans and through repayment guarantees on loans made
by other lenders.  Since the 1930s, the federal government has
provided billions of dollars in direct electricity loans and
guarantees on loans made by other lenders primarily to cooperatives
that serve rural areas. 


--------------------
\1 GAO/AIMD-97-110. 


   RUS' ELECTRICITY LOAN PROGRAMS
-------------------------------------------------------- Appendix II:1

Established by the Federal Crop Insurance Reform and the Department
of Agriculture Reorganization Act of 1994, RUS administers the
electricity programs that were operated by the former Rural
Electrification Administration (REA).\2 The Congress created REA in
1935 as part of a coordinated federal effort intended not only to
improve living conditions in rural areas, but also to alleviate the
high unemployment the nation experienced during the Depression. 
Because of higher construction and servicing costs, investor-owned
electric utilities had not extended service to many sparsely
populated areas of the country.  To fulfill its mission, REA
developed loan programs to assist rural areas in building and
operating electric generating facilities as well as wholesale
transmission and local distribution lines.  REA provided credit
assistance primarily to cooperatives owned by the consumers.  These
programs have been successful in helping farms and rural households
gain access to electrical service.  In 1940, about 25 percent of all
households in the nation were without electricity, but about 70
percent of farms did not have electrical service.  Today, virtually
all households are electrified. 

RUS makes direct loans primarily to construct and maintain
electricity distribution facilities that provide electricity to rural
areas.  RUS makes direct loans at below-market interest rates
according to law.  For these loans, it receives annual appropriations
to cover the interest differential.  RUS offers direct loans with a 5
percent interest rate to borrowers that serve financially distressed
rural areas, as well as municipal rate loans with a maximum 7 percent
interest rate to borrowers that meet certain criteria.  RUS also
provides 100 percent repayment guarantees on loans made by the
Federal Financing Bank and commercial lenders to finance the
construction, repair, and improvement of electricity generating and
transmission assets. 

RUS' electricity loans are made primarily to rural electric
cooperatives; more than 99 percent of the borrowers with electricity
loans are nonprofit cooperatives.  These cooperatives are either
generation and transmission (G&T) cooperatives or distribution
cooperatives.  A G&T cooperative is a nonprofit rural electric system
whose chief function is to sell electric power on a wholesale basis
to its owners, which consist of distribution cooperatives and other
G&T cooperatives.  A distribution cooperative sells the electricity
it buys from a G&T cooperative to its owners, the retail customers. 
In September 1997, we reported that RUS had 55 G&T borrowers and 782
distribution borrowers located throughout the country with
outstanding electricity loans.\3

Although operating somewhat like a commercial lender for rural
utilities, RUS is not required or intended to recover all of its
financing or other costs.  RUS' primary function is to provide credit
assistance to aid in rural development.  Interest charges to its
borrowers cover only a portion of the federal government's cost for
RUS' electricity loan programs. 


--------------------
\2 RUS also administers the former REA's telecommunications programs
and the water and waste disposal programs that were operated by the
former Rural Development Administration.  In this report, we discuss
only the electricity segment of RUS' loan programs. 

\3 GAO/AIMD-97-110. 


   DIRECT LOANS RESULTED IN NET
   FINANCING COSTS TO THE FEDERAL
   GOVERNMENT
-------------------------------------------------------- Appendix II:2

In a September 1997 report,\4 we estimated that RUS' net financing
cost to the federal government for its electricity loan program
totaled about $3.8 billion (in constant 1996 dollars) cumulatively in
fiscal years 1992 through 1996.  This net financing cost exists
because the annual interest income received from RUS borrowers is
substantially less than the federal government's annual interest
expense on funds provided to borrowers.  In addition, interest income
is affected by favorable rates and terms given to some borrowers and
also by financially troubled RUS borrowers that have missed scheduled
loan payments.  For example, one G&T borrower has not been required
to make interest payments on its $4.2 billion debt since filing for
bankruptcy in December 1994.  Meanwhile, the federal government
continues to incur interest expense on financing related to this
borrower. 

In April 1997, we reported that during fiscal years 1992 through
1996, RUS made or provided guarantees on 880 electricity loans, which
totaled about $4.35 billion.  Direct loans accounted for 835 of the
total number of loans and for about $3.3 billion of the total amount
of loans.  The other 45 electricity loans had RUS guarantees.  About
59 percent of the electricity loans were direct loans made at a 5
percent interest rate; these loans accounted for about 42 percent of
the total dollar amount of all electricity loans.\5

Until the Congress amended the Rural Electrification Act in 1973,
almost all financing was through direct loans from REA to electric
borrowers at a fixed rate of 2 percent with maturities up to 35
years.  The 1973 amendment increased the interest rate on the direct
loans from 2 percent to 5 percent.  The Congress amended the act
again in 1993 to provide direct loans with an interest rate that is
(1) tied to an index of municipal borrowing rates or (2) fixed at 5
percent.  Most loans are now made at the municipal rate with or
without a 7-percent cap.  Certain borrowers with customers that have
low consumer and household incomes and high residential retail rates
qualify for a loan at the 5 percent hardship interest rate. 


--------------------
\4 Federal Electricity Activities:  Appendixes to the Federal
Government's Net Cost and Potential for Future Losses, Volume 2
(GAO/AIMD-97-110A, Sept.  19, 1997). 

\5 Rural Development:  Financial Condition of the Rural Utilities
Service's Loan Portfolio (GAO/RCED-97-82, Apr.  11, 1997). 


   RUS' OUTSTANDING LOANS ARE OWED
   BY BORROWERS WITH FAVORABLE
   FINANCIAL CHARACTERISTICS
-------------------------------------------------------- Appendix II:3

In our April 1997 report,\6 we found that a majority of electricity
borrowers had generally favorable financial characteristics at the
end of calendar year 1995.  For example, we found that 804
distribution borrowers had average assets of $37.4 million,
liabilities of $21.6 million, and a net worth of $15.8 million.  Only
two of these borrowers had a negative net worth, and these two
borrowers owed about $32 million on their outstanding loans as of
September 30, 1996.  Of 51 power supply borrowers with outstanding
electricity loans at the end of 1995, 8 had a negative net worth. 
Seven of these eight borrowers owed about $6.1 billion on their
outstanding electricity loans as of September 30, 1996.\7

Most of the borrowers also had a net income at the end of 1995.  All
but 34, or 4.2 percent, of the electricity distribution borrowers had
a net income in 1995.  The 34 borrowers that had a loss owed $359
million on their outstanding electricity loans as of September 30,
1996.  Furthermore, 10 of these 34 borrowers had losses in at least 1
year between 1992 and 1994.  Only four of the power supplier
borrowers did not have a net income in 1995.  These four borrowers
owed $866 million for their outstanding electricity loans as of
September 30, 1996.  In addition, two of these four borrowers had
losses in at least 1 year between 1992 and 1994. 


--------------------
\6 GAO/RCED-97-82. 

\7 The electricity loans of the eighth borrower were settled on
September 13, 1996, when the borrower made a partial payment and RUS
wrote off the remaining debt. 


   ABOUT ONE-THIRD OF OUTSTANDING
   LOAN DEBT WAS OWED BY BORROWERS
   WITH FINANCIAL PROBLEMS
-------------------------------------------------------- Appendix II:4

As of September 30, 1996, RUS' borrowers owed about $32.3 billion in
outstanding debt on RUS' electricity loans.  As we reported in
September 1997, about $10.5 billion of the $32.3 billion was owed by
13 financially stressed borrowers.  Borrowers considered financially
stressed have either defaulted on their loans, had their loans
restructured but are still experiencing financial difficulty, filed
for bankruptcy, or have formally requested financial assistance from
RUS.  Of these 13 financially stressed borrowers, 4 borrowers are in
bankruptcy and have a total of about $7 billion in outstanding debt. 
The remaining nine borrowers have investments in uneconomical
generating plants and/or have requested financial assistance in the
form of debt forgiveness from RUS.  According to RUS officials, these
plant investments became uneconomical because of cost overruns,
continuing changes in regulations, and soaring interest rates.  These
investments resulted in high levels of debt and debt-servicing
requirements, making power produced from these plants expensive. 
Most of the electricity loans to RUS' problem borrowers were made
many years ago--some dating back to the 1970s. 


   SUBSTANTIAL LOAN WRITE-OFFS
   OCCURRED IN RECENT YEARS
-------------------------------------------------------- Appendix II:5

During fiscal years 1996 and 1997, RUS wrote off about $1.6 billion
in loans to rural cooperatives.  In our September 1997 report, we
reported that in fiscal year 1996, one G&T borrower made a lump sum
payment of $237 million to RUS in exchange for RUS writing off and
forgiving the remaining $982 million of its loan balance.  This
borrower's financial problems stemmed from its participation in a
nuclear plant construction project that experienced lengthy delays as
well as severe cost escalation.  When construction of the plant began
in 1976, its total cost was projected to be $430 million.  However,
according to the Congressional Research Service, the accrued
expenditures by 1988 were $3.9 billion as measured in nominal terms
(1987 dollars).  These cost increases are primarily the result of
changes in the Nuclear Regulatory Commission's health and safety
regulations after the Three Mile Island accident.  The remaining
increases are generally the result of inflation over time and
capitalization of interest during the delays. 

In the early part of fiscal year 1997, another G&T borrower made a
lump sum payment of about $238.5 million in exchange for forgiveness
of its remaining $502 million loan balance.  The G&T borrower and its
six distributor cooperatives borrowed the $238.5 million from a
private lender, the National Rural Utilities Cooperative Finance
Corporation.  The G&T borrower had originally borrowed from RUS to
build a two-unit coal-fired generating plant and to finance a coal
mine that would supply fuel for the generating plant.  The plant was
built in anticipation of industrial development from the emerging
shale oil industry.  However, the growth in demand did not
materialize, and there was no market for the power.  Although the
borrower had its debt restructured in 1989, it still experienced
financial difficulties as a result of a depressed power market.  RUS
and the Department of Justice decided that the best way to resolve
the matter was to accept a partial lump sum payment on the debt
rather than force the borrower into bankruptcy. 


   ADDITIONAL LOSSES FROM
   ELECTRICITY LOANS MAY OCCUR IN
   THE FUTURE
-------------------------------------------------------- Appendix II:6

In addition to the financially stressed loans, RUS has loans
outstanding to G&T borrowers that are currently considered viable by
RUS but may become stressed in the future because of high costs and
competitive or regulatory pressures.  We believe that some losses to
the federal government from currently viable loans are probable in
the future.\8 We believe the future viability of these G&T loans will
be determined on the basis of the borrower's ability to be
competitive in a deregulated market.  For example, 27 of 33 loans to
G&T borrowers had high average production revenues in comparison to
regional investor-owned utilities, and 17 of the 33 had higher
average production revenues than publicly owned utilities.  The
relatively high average production costs indicate that the majority
of G&T borrowers may have difficulty competing in a deregulated
market.  According to RUS, several borrowers had already requested
forgiveness or a restructuring of their debt because they did not
expect to be competitive because of high costs.  However, RUS
officials stated that they will not write off debt solely to make
borrowers more competitive. 


--------------------
\8 We based our discussion of the risk of nonrecovery on Statement of
Federal Financial Accounting Standards No.  5, Accounting for
Liabilities of the Federal Government, which indicates that if the
chance that a contingent loss will occur is more likely than not,
then the loss is considered "probable"; if the chance is more than
remote but less than probable, then the loss is considered
"reasonably possible"; and if the chance is slight, then the loss is
considered "remote."


GAO'S PRIOR WORK ON THE BONNEVILLE
POWER ADMINISTRATION
========================================================= Appendix III

The Bonneville Power Administration (Bonneville), the largest of the
power marketing administrations (PMA) in terms of generating capacity
and sales, has been a low-cost supplier of electricity.  In September
1997, however, we noted that its power has lost some of its price
advantage, as a result of such factors as low prices for natural gas
(the fuel used by Bonneville's competitors to generate low-cost
power), surplus generating capacity on the West Coast, the opening of
the competitive wholesale electricity market, and the resulting
decline in electricity prices.  It has also had higher costs because
of requirements for fish recovery, resource acquisitions, and other
factors.  Bonneville's ability to reduce costs is hampered by the
fact that a large part of its costs are fixed.  The ultimate risk,
should Bonneville be unable to cover its costs, will be the
Treasury's.\1


--------------------
\1 GAO/AIMD-97-110. 


   BONNEVILLE MARKETS POWER IN THE
   PACIFIC NORTHWEST
------------------------------------------------------- Appendix III:1

Bonneville was created in 1937 by the Bonneville Project Act,
originally as an interim agency to market electric power produced by
the Bonneville Dam, then under construction on the Columbia River. 
In 1940, Bonneville's marketing responsibilities were broadened to
include the power from Grand Coulee Dam in central Washington. 
Today, Bonneville markets electric power from the Federal Columbia
River Power System, which consists of 29 federally owned
hydroelectric projects, most of which are in the Columbia River
Basin, and one nonfederal nuclear plant of the Washington Public
Power Supply System.  The Federal Columbia River Power System
provides roughly half the power used in the Pacific Northwest. 
Bonneville, the Corps of Engineers, and the Bureau of Reclamation
coordinate the system's operation with many public and privately
owned utilities that own dams on the river system. 

Like other PMAs, Bonneville sells primarily wholesale power from the
dams and other generating plants to public and private utilities and
direct service industries.  By law, it gives preference to public
utilities and sells excess power only outside of its primary customer
service area--300,000 square miles in the Pacific Northwest, made up
of Idaho, Oregon, Washington, western Montana, and small portions of
California, Nevada, Utah, and Wyoming.\2

Bonneville builds, owns, and operates over 15,000 miles of
transmission lines that make up 75 percent of the Northwest's
high-voltage transmission capacity.  Over the years, the Congress has
expanded Bonneville's mission to include conservation and renewable
resource development, rate relief for specified residential and small
farm power users, and specific mandates for fish and wildlife
protection and funding. 


--------------------
\2 In December 1997, Bonneville announced that it would begin selling
power to its first preference customer in eastern Montana. 


   BONNEVILLE'S POWER PROGRAM IS
   TO BE SELF-SUPPORTING
------------------------------------------------------- Appendix III:2

Unlike the other PMAs, Bonneville no longer receives an annual
appropriation from the Congress.  The Federal Columbia River
Transmission System Act of 1974 placed Bonneville on a self-financing
basis--its operating expenses are to be paid for by revenues from the
sale of power and transmission service.  Funds received from
customers are paid to Bonneville, which then deposits the receipts
into a special Bonneville fund at the Treasury.  Expenditures for
Bonneville are paid from that special fund.  For capital
expenditures, Bonneville has the authority to borrow from the
Treasury.  Its Treasury bond borrowing authority is capped at $3.75
billion ($2.5 billion for transmission and other capital investments
and $1.25 billion for conservation and renewable energy investments). 
Bonneville is required to set its rates for power and transmission
sales at levels that generate revenues sufficient to cover annual
expenses and pay back previously appropriated funds.  Bonneville is
required to make an annual payment to Treasury that includes
debt-servicing costs on appropriated debt and Treasury bonds. 
Similar to the other PMAs, Bonneville is also required to recover and
repay to the Treasury the operating agencies' power-related capital
and operating expenses. 


   BONNEVILLE'S DEBT EXCEEDS $17
   BILLION
------------------------------------------------------- Appendix III:3

Unlike the other PMAs, Bonneville has a legislative mandate that
requires it, within certain limits, to provide sufficient firm power
to meet the needs of the customers in its primary service area. 
Because of this mandate, and in response to its estimate of growing
energy demand in the Pacific Northwest, Bonneville entered into
nonfederal financing agreements to acquire all or part of the
capability of four nuclear power plants constructed, owned, and to be
operated by other entities.  As part of these agreements, Bonneville
was required to pay the projects' annual costs, including debt
service, in amounts ranging from 30 to 100 percent of total costs
incurred.  Later, a variety of events, including construction cost
overruns and overly optimistic estimates of electricity demand, made
it clear that some of these plants would not be economical to
complete or operate.  Accordingly, construction was halted on two of
these nuclear plants and they were not completed.  In addition, one
previously operating plant has been shut down permanently.  As a
result, Bonneville is responsible for about $4.2 billion in
nonfederal debt associated with three nonoperating nuclear plants and
an additional $2.5 billion in nonfederal debt associated with the one
operating nuclear plant.\3 Bonneville's total debt exceeded $17
billion, as of September 30, 1996. 


--------------------
\3 The nonfederal debt also consists of $321 million invested in
small hydroelectric projects and conservation measures. 


   RISK OF LOSS FROM BONNEVILLE IS
   REMOTE THROUGH FISCAL YEAR 2001
   BUT INCREASES THEREAFTER
------------------------------------------------------- Appendix III:4

Bonneville's high fixed costs limited its ability to respond to
competition by decreasing rates and contributed to a loss of
customers in recent years.  Although we concluded in a September 1997
report that the risk of any significant loss to the federal
government from Bonneville is remote through fiscal year 2001,\4
thereafter, the expiration of customer contracts, risks from market
uncertainties, Bonneville's high fixed costs, and upward pressure on
operating expenses increase the risk of loss to the federal
government.\5

Despite a number of factors that mitigate this risk, we reported that
it is reasonably possible that the federal government will incur
losses from Bonneville after fiscal year 2001.  In addition, one
small project that serves Bonneville represents a probable loss to
the federal government. 


--------------------
\4 GAO/AIMD-97-110. 

\5 We based our discussion of the risk of nonrecovery on Statement of
Federal Financial Accounting Standards No.  5, Accounting for
Liabilities of the Federal Government, which indicates that if the
chance that a contingent loss will occur is more likely than not,
then the loss is considered "probable"; if the chance is more than
remote but less than probable, then the loss is considered
"reasonably possible"; and if the chance is slight, then the loss is
considered "remote."


      KEY FACTORS STABILIZE
      BONNEVILLE THROUGH FISCAL
      YEAR 2001
----------------------------------------------------- Appendix III:4.1

Three key factors have stabilized the government's risk of loss
attributable to Bonneville through fiscal year 2001 and, in our view,
make risk remote for this period.  First, in 1995 and 1996,
Bonneville signed its customers to contracts to purchase a
substantial amount of power through fiscal year 2001.  Bonneville
projects that firm power sales to these customers will secure about
$1.14 billion annually through fiscal year 2001, or 63 percent of
each year's total projected power revenues.  Second, Bonneville's
management entered into a memorandum of agreement with various
federal agencies that has limited its fish mitigation costs through
fiscal year 2001.  This agreement also created a contingency fund of
$325 million for Bonneville's past nonpower fish mitigation
expenditures.\6 Finally, Bonneville had strong water years in 1996
and in 1997 and estimates that it will have a financial reserve of
about $400 million at the end of fiscal year 1997.\7 In addition, the
$325 million fish cost contingency fund is available under specified
circumstances. 


--------------------
\6 The Northwest Power Act requires Bonneville to protect, mitigate,
and enhance fish and wildlife resources to the extent these resources
are affected by federal hydroelectric projects.  The act also directs
Bonneville to allocate fish and wildlife costs to the projects'
various purposes, for example, flood control, irrigation, and power. 
The reserve represents the portion of Bonneville's expenditures that
are related to nonpower uses of the projects.  To the extent
Bonneville uses the $325 million reserve, the federal government will
incur these costs because the memorandum of agreement allows it to
apply the $325 million, under specified circumstances, as a credit
against its Treasury payment. 

\7 Bonneville's financial reserves of about $400 million include cash
and deferred Treasury borrowing authority.  Deferred borrowing
authority is created when Bonneville uses operating revenues to
finance capital expenditures in lieu of borrowing.  This temporary
use of cash on hand instead of borrowed funds creates the ability in
future years to borrow money, when fiscally prudent, to liquidate
revenue-funded activities.  The deferred Treasury borrowing authority
is similar to an unused line of credit.  While this may be useful in
the short term to provide liquidity, its use results in additional
debt; thus, deferred borrowing authority is not a long-term solution
to financial difficulty. 


      RISK INCREASES AFTER FISCAL
      YEAR 2001
----------------------------------------------------- Appendix III:4.2

After fiscal year 2001, Bonneville faces the expiration of customer
contracts, significant market uncertainties, high fixed costs, and
significant upward pressure on operating expenses.  Nearly all of
Bonneville's power contracts with customers expire at the end of
fiscal year 2001.  If these customers can find power cheaper than
Bonneville can offer, they may opt to leave Bonneville.  One of the
key market uncertainties that will determine whether cheaper power
will be available is the future production cost of gas-fired
generation plants.  This generation source has become increasingly
competitive because of low natural gas prices and improving gas
turbine technology.  Natural gas prices in the Pacific Northwest are
low as the result of several factors, including a large supply coming
from Canada.  Also, recent technology advances have improved the
efficiency of gas turbines by more than 50 percent.  According to
Bonneville, natural gas-generated power has driven down the price of
wholesale electricity and resulted in customers leaving or obtaining
some of their power at rates well below Bonneville's current rate. 

According to Bonneville, a surplus of power on the West Coast is also
driving down the price of wholesale power.  Because utilities still
are able to pass on fixed costs to captive retail customers, surplus
wholesale power is being sold on a marginal cost basis.  According to
Bonneville, other utilities and power marketers\8 are offering
wholesale power as low as 1.5 cents per kilowatthour (kWh), which is
lower than Bonneville's 2.14 cents per kWh for sales of comparable
products.  However, it is uncertain whether surplus power and
low-cost natural gas generation will continue to drive down wholesale
power prices after fiscal year 2001. 

It is also uncertain what impact retail open access will have on
Bonneville's competitive position.  Retail open access--which would
provide retail consumers the freedom to choose among suppliers--could
result in Bonneville's wholesale customers being uncertain about the
size of their own future power needs.  These power needs will be
directly affected by retail customers' choices about their suppliers. 
Bonneville's customers may be hesitant to sign long-term contracts to
purchase power from Bonneville to the extent that they face
uncertainty about future power needs.  However, even without
long-term contracts, Bonneville is likely to remain a major supplier. 

Bonneville's substantial fixed costs will continue to inhibit its
flexibility to lower its rates and meet competitive pressures.  For
example, 32 percent of Bonneville's revenue went to pay financing
costs in fiscal year 1996--substantially more than a nationwide
average of 14 percent for IOUs and 18 percent for publicly owned
generating utilities.  After fiscal year 2001, Bonneville will
continue to face high fixed costs relating to its $17 billion debt. 

Bonneville will also face significant upward pressure on its
operating expenses after fiscal year 2001.  The most significant of
these operating expenses is fish mitigation.  It is uncertain whether
an agreement similar to the current agreement will be possible after
the present one expires.  Without this agreement, Bonneville is at
risk of escalating costs after fiscal year 2001 if additional funds
for fish measures beyond those planned at this time are needed.\9
Bonneville also faces new or additional costs after 2001.  First, it
plans to implement a phased-in approach to recovering the full cost
of pension and postretirement health benefits in fiscal year 1998 but
will defer full recovery until fiscal year 2002, when $55 million
will be due.  To completely recover obligations for fiscal years 1998
through 2001, an additional $35 million will be due in fiscal year
2003.  Other new or additional costs that will be incurred after
fiscal year 2001 include $806 million in irrigation debt payments and
$396 million in payments to the Confederated Tribes of the Colville
Reservation for the tribes' share of the Grand Coulee Dam revenues. 
These costs would be paid out over several decades. 


--------------------
\8 Power marketers are subsidiaries of IOUs or independent companies
that buy and sell power, typically on a wholesale basis. 

\9 If total federal mitigation costs increase and Bonneville reduces
or caps its fish mitigation expenses after 2001, the federal
government may have to bear additional costs. 


      MITIGATING FACTORS REDUCE
      PROBABILITY OF LOSS
----------------------------------------------------- Appendix III:4.3

Several factors mitigate the federal government's risk of future
losses relative to Bonneville.  These factors include certain
inherent cost advantages, management actions to reduce operating
costs, and an extensive transmission system.  We believe that these
factors reduce the risk of loss to the federal government after 2001
but that the risk is still reasonably possible.  Moreover, Bonneville
is scheduled to have nearly all of its nonfederal debt paid off by
2019, with a substantial decrease in debt service beginning in 2013. 
If Bonneville is able to make these payments as scheduled, all else
being equal, its fixed financing costs would be more in line with
those of its competitors.  This would reduce the risk to the federal
government.  As shown in figure III.1, Bonneville's 1995 average
revenue per kWh was more than 15 percent lower than the average
revenues of IOUs and publicly owned generating utilities in the
primary North American Electric Reliability Council\10 region
(Western Systems Coordinating Council) in which Bonneville operates. 

   Figure III.1:  Average Revenue
   per kWh for Wholesale Power
   Sold in 1995 for Bonneville,
   IOUs, and Publicly Owned
   Generating Utilities

   (See figure in printed
   edition.)

Legend:  Bonneville - Bonneville Power Administration; IOU -
Investor-owned utility

Note:  The latest data available for IOUs and publicly owned
generators were for 1995.  We included Bonneville's 1996 average
revenue per kWh to show that it decreased almost 20 percent from 1995
to 1996. 

Source:  GAO's analysis of Bonneville's annual reports, preliminary
(unaudited) 1995 IOU data from the Energy Information Administration,
and publicly owned generators' data from the American Public Power
Association. 

As previously mentioned, Bonneville is facing significant
competition.  However, its management believes that its average
production costs are less than those of others in the Pacific
Northwest, as shown in figure III.1.  If the supply of surplus power
dwindles and gas generation costs rise, which Bonneville believes
will happen, Bonneville's low average production costs should improve
its long-term competitive position.  This long-term position will be
further improved after 2012 if Bonneville repays its nonfederal debt
as scheduled. 

Bonneville has comparatively low average production costs because of
certain inherent cost advantages it has over nonfederal utilities. 
For example, in 1996 Bonneville did not recover nearly $400 million
of the costs associated with producing and marketing federal power. 
In addition, the hydroelectric plants that generate the power
marketed by all the PMAs have cost advantages over coal and nuclear
generating plants, which generate over 81 percent of the electricity
in the United States.  Bonneville's hydroelectric plants, which were
built decades ago, also had relatively low construction costs
compared with newer, nonfederal utilities' construction.  Other
advantages are that Bonneville, like the other PMAs, generally does
not pay taxes, and the interest income that bondholders receive from
Bonneville's nonfederal debt is exempt from the federal personal
income tax and some state income taxes. 

Bonneville's management has taken significant steps in the last
several years to respond to the intense wholesale electricity
competition in the Pacific Northwest.  According to Bonneville, its
staff decreased from about 3,755 in March 1994 to 3,160 by the end of
fiscal year 1996.  An additional reduction to 2,755 is planned by
fiscal year 1999.  In addition, over the last several years,
Bonneville has refinanced much of its Treasury bonds and nonfederal
debt to keep its interest expense as low as possible.  According to
Bonneville, these staffing and other cost savings will reduce planned
expenses by an average of $600 million per year during fiscal years
1997 through 2001 and have allowed a 13-percent rate decrease for
those years. 

Bonneville also has an extensive transmission system that constitutes
about 75 percent of the bulk power transmission capacity in the
Pacific Northwest.  According to Bonneville, if it is unable to sell
its power at a level that recovers all costs, it may be able to use
revenues from the sale of transmission services to help recover
stranded costs.\11 This could involve allocating stranded generation
costs, in whole or in part, to transmission charges. 


--------------------
\10 The North American Electric Reliability Council was formed by the
electric utility industry to promote the reliability and adequacy of
the bulk power supply in the electric utility systems of North
America.  The Council consists of 10 regional reliability councils
and encompasses essentially all the power systems of the contiguous
United States, as well as parts of Canada and Mexico. 

\11 As defined by FERC, a stranded cost is any legitimate, prudent,
and verifiable cost incurred by a public or transmitting utility that
is no longer economically viable in a competitive wholesale
environment. 


   RISK OF LOSS FROM TETON DAM
   PROJECT IS PROBABLE
------------------------------------------------------- Appendix III:5

We identified one small Bonneville project where the loss to the
federal government is probable.  Teton Dam was a multipurpose project
built by the Bureau of Reclamation on the Teton River in Idaho.  The
dam failed in 1976 when it was substantially complete, resulting in
flooding, loss of life, and loss of the facilities.  Had the project
been completed, power-related construction costs of about $7.3
million and irrigation costs of about $56.6 million would have been
included in Bonneville's power rates for eventual repayment to the
federal Treasury. 

Since the failure of the dam in 1976, the project's costs have been
carried on the books of the Bureau as construction work-in-progress. 
While assets of this type normally accrue interest charges, the Teton
project has accrued no interest since 1976.  Since that time,
interest charges of about $5 million, at the project's interest rate
of 3.25 percent, would normally have been paid to the Treasury, as we
reported in September 1997.\12

The project's power-related construction costs are part of
Bonneville's appropriated debt balance.  However, provisions to
recover this amount have not been made.  According to Bonneville,
since the project was not formally completed and placed in service,
its costs cannot be put into Bonneville's rates.  According to the
Bureau, it has no plans for further construction at the site and the
project should be written off; however, according to the Bureau, a
write-off would require deauthorization of the project by the
Congress.  Whether or not the project is deauthorized, we believe
these costs are unlikely to ever be recovered. 


--------------------
\12 GAO/AIMD-97-110A. 


OBJECTIVES, SCOPE, AND METHODOLOGY
========================================================== Appendix IV

From the early 1900s through September 30, 1996, federal agencies
that generate and/or market electricity and that make or guarantee
loans to finance improvements to electricity systems incurred a debt
of about $84 billion.\1 Like the other federal agencies, the
Southeastern, Southwestern, and Western Area power
administrations--responsible for about $7 billion of this debt--face
an uncertain future as electricity markets restructure.  In response,
the Chairmen of the House Committee on Resources and the Subcommittee
on Water and Power asked us to focus on these three power marketing
administrations (PMA) and (1) to examine whether the government
operates them and the related electric power assets in a businesslike
manner that recovers the federal government's capital investment in
those assets and the costs of operating and maintaining them and (2)
identify options that the Congress and other policymakers can pursue
to address concerns about the role of the three PMAs in emerging,
restructured markets or to manage them in a more businesslike
fashion.  Our options also have implications for the Army's Corps of
Engineers (Corps) and the Department of the Interior's Bureau of
Reclamation (Bureau), which generate most of the power these PMAs
market.  As requested, the report also provides information on the
Tennessee Valley Authority (TVA), Rural Utilities Service, and
Bonneville Power Administration (Bonneville), which is contained in
appendixes I, II, and III, respectively. 

We also included in this report information from our more generalized
reports that address topics concerning the ways that federal agencies
can be operated in a more businesslike fashion.  See Related GAO
Products at the end of this report for a list of the products we
used. 


--------------------
\1 GAO/AIMD-97-110. 


   EXAMINING WHETHER THE
   GOVERNMENT OPERATES ITS POWER
   ASSETS TO RECOVER COSTS AND
   PROMOTE REPAYMENT OF THE
   FEDERAL INVESTMENT
-------------------------------------------------------- Appendix IV:1

To examine whether the federal government operates its electric power
and related assets in a manner that recovers the associated costs and
promotes the repayment of the federal investment in those assets, we
first researched the history of the nation's electric power industry,
focusing on the evolution of markets and regulatory structures.  Our
work included reviewing the effects of major statutes and their
amendments, such as the Federal Power Act of 1920, the Public Utility
Holding Company Act of 1935, the Public Utility Regulatory Policies
Act of 1978, and the Energy Policy Act of 1992.  In addition, we
examined the roles of investor-owned utilities; cooperatives;
publicly owned, nonfederal utilities (that is, those owned by state,
municipal, or other nonfederal public entities); federal generators
and marketers of power utilities (including the PMAs, the operating
agencies, and TVA), and RUS.  We monitored current changes in the
industry, especially those pertaining to restructuring and retail
competition, by contacting associations of electric power providers
in Washington, D.C.  (the American Public Power Association, the
Edison Electric Institute, the National Hydropower Association, and
the National Rural Electric Cooperatives Association), DOE, the PMAs,
and the Federal Energy Regulatory Commission (FERC) and by reviewing
state public utility commission homepages on the Internet and
industry publications. 

In addition, we reviewed our recent products on the business
practices of the PMAs and the operating agencies, including (1)
whether the PMAs' rates recover all of the costs associated with
generating, transmitting, and marketing electricity and (2) the
related costs that are assigned to power for repayment, such as
assistance to irrigation, and the rate and repayment methodologies of
the PMAs.  To the extent deemed appropriate, we followed up on issues
from our prior work with field work at various locations of
Southeastern, Southwestern, Western, the Bureau, the Corps, and
various power customer groups (namely, the Southeastern Federal Power
Customers, Atlanta, Georgia, and the Midwest Electric Consumers
Association, Denver, Colorado).  For example, at the Billings office
of the Bureau, we updated previous information about the Bureau's
efforts to recover over $450 million in federal investment in
hydropower capacity and reservoir storage for planned irrigation
projects. 


   IDENTIFYING OPTIONS FOR THE
   THREE PMAS AND THEIR OPERATING
   AGENCIES
-------------------------------------------------------- Appendix IV:2

We identified options that the Congress and other policymakers can
pursue to address concerns about the role of these three PMAs in
restructuring markets or to manage them in a more businesslike
fashion.  To identify these options, we consulted officials from the
American Public Power Association, the Edison Electric Institute, the
National Hydropower Association, the National Rural Electric
Cooperatives Association, and the Office of Management and Budget, in
Washington, D.C.  In addition, we contacted the Bureau, the Corps'
Hydropower Coordinator, and DOE's Power Marketing Liaison Office (on
behalf of Southeastern, Southwestern, and Western), and the
Department of the Interior.  We also contacted the Bureau's offices
in Billings, Montana; Denver, Colorado; Sacramento, California; and
Salt Lake City, Utah.  We discussed options with representatives of
Southeastern in Elberton, Georgia; of Southwestern in Tulsa,
Oklahoma; and of Western in Billings, Montana; Golden, Colorado; and
Salt Lake City, Utah.  We also discussed options with or obtained
information from the PMAs' preference customers or customer groups,
such as the Midwest Electric Consumers Association and Western States
Power Corporation, Denver, Colorado; the Southeastern Federal Power
Customers, Inc., Atlanta, Georgia; and the Southwestern Power
Resources Association, Tulsa, Oklahoma; and, in some cases, their
legal counsels. 

A primary task in examining the option to divest the PMAs was to
estimate the effects of a divestiture on the rates paid by the PMAs'
customers.  In this connection, we estimated how much the PMAs'
existing customers' rates might change if the PMAs were sold.  To
calculate these changes, we compared (1) the average blended rate
that each PMA customer paid for wholesale power from all sources in
1995 with (2) the wholesale rate that each PMA customer might pay
after divestiture.  The difference in these two rates equals the
change in rates attributable to a divestiture. 

Estimating the potential rate changes required several steps and
assumptions.  First, we estimated the average rate that each PMA
customer paid for PMA and non-PMA power in 1995.  To calculate how
much customers paid for the PMAs' power, we obtained data from
Southeastern's, Southwestern's, and Western's fiscal year 1995 annual
reports.  Then, to learn how much each PMA customer paid for the
wholesale power it purchased from other sources, we used the sales
for resale databases compiled by DOE's Energy Information
Administration (EIA).\2 We found that for about one-third of the
PMAs' total customers, EIA's data lacked the volumes of wholesale
power the customers purchased from non-PMA sources, the amount the
customer paid for the power, or both.\3 In these cases, we assumed
the customer paid a rate equal to the average market rate paid by
customers of the same type (for example, municipal utilities and
cooperatives) for wholesale power in the customer's state.  We then
blended each customer's PMA and non-PMA purchases to estimate how
much the customer paid for wholesale power from all sources in 1995. 

Second, to estimate how much each PMA customer would pay for power
after a divestiture, we assumed each PMA customer would pay a rate
that equals the average rate it paid for wholesale power from sources
other than the PMAs in 1995.  We used this assumption because it is
likely that in the period immediately after a divestiture, the new
owners of the PMAs' assets would charge the prevailing market rates
for wholesale power in the area.  We took this approach because we
were unable to obtain forecasts of future wholesale rates.  Although
EIA recently used its National Energy Modelling System to forecast
future electricity rates,\4 according to the agency, its projections
are only for retail rates.\5 Other projections of future wholesale
rates were proprietary. 

Finally, after calculating how much each PMA customer paid for PMA
power in 1995 and how much it would pay for PMA power after a
divestiture, we calculated the difference (both in percentage and
cents per kWh) between the two rates.  These differences represent
our estimates of each customer's potential increase in average
blended rates following a divestiture of the PMAs. 

It is important to note that because we assume, after divestiture,
that each customer will pay a rate for power that equals what the
customer paid for non-PMA power in 1995, our methodology is
conservative.  If prices for wholesale power decline in the future,
as many industry analysts believe they will, each customer's change
in rates from divestiture of the PMAs will be smaller than our
estimates. 

To estimate how each preference customer's rate change would affect
the rates paid by its residential end-users, we assumed that (1) each
preference customer would pass all the rate increase from divestiture
onto its end-users and (2) that residential end-users consume 10,037
kWh of electricity per year.  The monthly increase in a residential
end-user's electricity bill equals the preference customer's rate
increase after divestiture (in cents per kWh) times residential
end-users' average annual electricity consumption (10,037 kWh),
divided by 12. 

We conducted our review from April 1997 through February 1998 in
accordance with generally accepted government auditing standards. 

We provided a draft of this report to the Department of Defense
(including the Corps); Bonneville; DOE's Power Marketing Liaison
Office that represented the views of Southeastern, Southwestern, and
Western; the Department of the Interior (including the Bureau); and
FERC.  Their comments and our responses are included in appendixes
VI, VII, VIII, IX, and X, respectively. 


--------------------
\2 Specifically, we used EIA's PURCH.Y95 and SALES.Y95 databases. 

\3 EIA officials stated that the data were missing for several
reasons, among them that the PMA customers involved were so small
they did not have to file the reports (FERC's Form 1, DOE's Form 412)
that EIA uses to compile the sales for resale data. 

\4 See Electricity Prices in a Competitive Environment (DOE/EIA-0614,
Aug.  1997). 

\5 We attempted to derive forecasts of wholesale prices from EIA's
retail price forecasts by subtracting distribution costs from EIA's
projections.  However, we found that our result was much higher than
the national average rate for wholesale power EIA reports in
Financial Statistics of Major U.S.  Investor-Owned Utilities.  After
consulting with EIA, we chose not to use its retail price forecasts
because they are based on EIA's judgmental assignment of electricity
generators' costs to services, such as generation, transmission, and
distribution, rather than actual sales data. 


PROJECTS AND RATE-SETTING SYSTEMS
OF THE THREE PMAS WHERE THE
FEDERAL INVESTMENT IS AT RISK
=========================================================== Appendix V

As shown in chapter 2, up to $1.4 billion in federal investment is at
various degrees of risk for nonrecovery at six of Southeastern's,
Southwestern's, and Western's projects and rate-setting systems.\1


--------------------
\1 We based our discussion of the risk of nonrecovery on Statement of
Federal Financial Accounting Standard No.  5, Accounting for
Liabilities of the Federal Government, which states that if the
chance that a contingent loss will occur is more likely than not, the
loss is to be described as "probable"; if the chance is more remote
but less than probable, it is "reasonably possible"; if the chance is
slight, it is "remote."


   RICHARD B.  RUSSELL PROJECT
--------------------------------------------------------- Appendix V:1

The Russell project, located on the Savannah River, which is the
border between Georgia and South Carolina, has four conventional
hydropower generating units (300 MW), which are operating, and four
pumping units (300 MW), which have not operated as intended.\2

Because of litigation over large fish kills, the pumping units, which
were completed in 1992, have not been allowed to operate
commercially.  As a result, the construction cost associated with
them has been excluded from power rates and is not being recovered. 
Moreover, the interest associated with these capital costs has not
been paid to the Treasury each year.  Instead, this
interest--estimated at about $29.9 million for fiscal year 1996--has
been capitalized and added to the construction-work-in-progress
balance each year.  As of September 30, 1996, we estimate that the
balance in the construction-work-in-progress account was about $518
million.  According to Southeastern's power customers, if the pumping
units become operational, then the construction costs would be
recovered through rates that, consequently, would increase by about
25 percent for customers of Southeastern's Georgia-Alabama-South
Carolina rate-setting system.  According to Southeastern's customers,
even with this increase, the system's rates would remain competitive. 
In our view, if the construction-work-in-progress costs are put into
the rates in the near future, then the risk of nonrecovery of the
$518 million remains.  However, the longer the delay in operating the
four pumping units, the greater the risk of nonrecovery because the
amount to be recovered will also increase.  At some point, the price
of the power for the Georgia-Alabama-South Carolina system may become
noncompetitive, and in such a situation, we believe the risk of some
loss to the federal government is reasonably possible.  If the
pumping units are never allowed to function, then it is probable that
the federal government will lose its entire $518 million investment. 
In commenting on our draft report, DOE's Power Marketing Liaison
Office noted that some unspecified portion of this investment will be
recovered even if the units are never commercially operated. 


--------------------
\2 During periods when the demand for power is low, pumping units
return water that has passed through the generating units to the
reservoir so that water can be reused to produce power during periods
when demand is higher. 


   HARRY S.  TRUMAN DAM AND
   RESERVOIR
--------------------------------------------------------- Appendix V:2

The Truman project, located on the Osage River in Missouri, has six
hydropower generating units (160 MW of nameplate capacity) placed in
service from 1980 to 1982 that are intended to act both
conventionally and as pumping units.  Because of design problems and
fish kills caused by the pumping capability, the generating units
have operated only as conventional units, not as pumping units.  Only
53 MW of generating capacity were declared to be operable. 
Consequently, it was determined that the costs associated with the
capacity that has not been allowed to operate commercially should not
be included in Southwestern's power rates.  Southwestern petitioned
FERC to defer recovery of these costs.  In 1989, FERC concurred with
Southwestern.  Thus $31 million is not being recovered through power
rates until the pumping units work as designed.  According to Corps
officials, three of the six units are now in service, operating as
conventional, not pumping, units.  Two more units were to be
rehabilitated and placed back on line by February 1998, and the last
unit is to be back on line by February 1999.  These last three units,
however, will also operate only in a conventional mode pending
lifting of an injunction by the State of Missouri.  Corps officials
stated that although the modifications should increase the
availability of the generating units, the fish kill issue has not
been resolved and associated capacity has not been restored as a
result.  Unless the pumping capacity becomes operational, which we
believe is unlikely given the amount of time it has been inoperable,
it is probable the government will lose the $31 million invested in
it.  If the units do come on line as designed, then the risk of
future losses is remote.  In commenting on our draft report, DOE's
Power Marketing Liaison Office noted that Southwestern can add to its
power repayment study the power-related costs of the pumpback units
even if the units are never operable. 


   CENTRAL VALLEY PROJECT
--------------------------------------------------------- Appendix V:3

California's Central Valley Project had an outstanding appropriated
debt of $267 million as of September 30, 1996, and its hydropower
program incurred a loss of $24 million in fiscal year 1996.  The
project has an installed generating capacity of about 2,000 MW at 12
hydropower plants.  Faced with competition from low-cost producers,
Western cut the project's power rates by 26 percent in fiscal year
1996.  As stated in chapter 1 of this report, Western also announced
a decrease of over 20 percent, effective October 1, 1997, in the
composite rates of power it markets from hydropower plants in the
Central Valley Project.  These rate cuts were facilitated, in part,
by renegotiating contracts that obligate Western to purchase power
for its customers if the project cannot supply enough.  We believe
that the extent to which any of Western's rate cuts will be
sustainable at competitive levels is unclear.  Moreover, the success
of Western in reestablishing and sustaining the competitiveness of
the project's power is uncertain because of environmental
legislation.  The Central Valley Project Improvement Act of 1992 adds
fish and wildlife mitigation, protection, and restoration as
authorized purposes for the project, thus restricting the use of
water for purposes such as hydropower generation, irrigation, and
municipal and industrial water.  These restrictions may reduce the
amount of power generated and make it uncertain whether revenues from
the sale of whatever amount of power that can be produced will repay
the federal investment in hydropower and other costs allocated for
repayment through power revenues.  For example, according to the
Bureau, an analysis of environmental impacts indicates that the
management of 800,000 acre-feet of water in the project for
environmental purposes may result in a reduction of about 5 percent
in hydropower production.  Moreover, according to Western officials,
when the reallocation of water required by the act occurs,
substantial nonpower costs may be reallocated to power for repayment,
thus placing further upward pressure on Western's power rates.  This
situation will reduce Western's ability to restore the
competitiveness of the project's power rates, according to Western
officials.  The 1984 Trinity River Basin Fish and Wildlife Management
Act also restricts the use of the project's water for generating
electricity.  These uncertainties, along with emerging competition,
lead us to conclude that it is reasonably possible that some of the
$267 million federal investment will not be repaid. 


   PICK-SLOAN MISSOURI BASIN
   PROGRAM
--------------------------------------------------------- Appendix V:4

The Pick-Sloan Missouri Basin Program is a comprehensive plan to
manage parts of 10 midwestern and western states that are drained by
the Missouri River.  The program's Eastern and Western divisions have
a total generating capacity of about 3,100 MW at 13 power plants.  In
May 1996, we estimated that about $454 million of the federal
investment in hydropower capacity initially designed for use by
future irrigation projects and in costs associated with storing water
for these projects would likely not be completed.\3 Although Western
has scheduled these costs for repayment through power revenues, this
will not occur until the future irrigation projects become operable. 
According to the Bureau, almost all of these planned irrigation
projects are infeasible and unlikely to be completed.  Under
applicable statutory repayment principles, recovery of these costs,
which we estimate at $464 million as of September 30, 1996, cannot
occur unless the associated irrigation projects come into service. 
Without legislative action, it is probable that Western will not be
required to recover the principal or any interest on the $464
million. 


--------------------
\3 Federal Power:  Recovery of Federal Investment in Hydropower
Facilities in the Pick-Sloan Program (GAO/T-RCED-96-142, May 2,
1996). 


   WASHOE PROJECT
--------------------------------------------------------- Appendix V:5

The Washoe Project with the associated Stampede Powerplant (10 MW),
located in east-central California and west-central Nevada, is not
generating sufficient revenue to cover its annual power-related
operating expenses, interest, or the federal investment in it.  Since
1988, deferred payments to the Treasury for its annual operating
expenses and interest charges totaled about $4.1 million through the
end of fiscal year 1996.  The project also had $8.9 million in
appropriated debt as of the end of fiscal year 1996.  To compound
matters, according to Western officials, the power plant would have
to price its power at a noncompetitive level--about 5.7 cents per
kWh, according to Western's estimates--to cover its operating
expenses (less depreciation), interest, and debt repayments.\4 To
recover the costs associated with the Washoe Project, Western
officials told us that they were considering combining the Washoe
Project's power with the power from the Central Valley Project and
establishing a blended rate.  However, because the Central Valley
Project itself faces challenges in remaining competitive, we
concluded that it is reasonably possible that the $13 million in
deferred interest and federal capital investment will not be
recovered.  The risk of nonrecovery worsens to probable if the Washoe
Project's power continues to be marketed on a stand-alone basis.  In
commenting on our draft report, DOE's Power Marketing Liaison Office
noted that Western staff are proposing the blending of the costs of
the power from the Washoe Project with those of Central Valley
Project after the year 2004. 


--------------------
\4 Power from the Washoe Project generated revenues of only 1.02
cents per kWh in fiscal year 1996. 


   MEAD-PHOENIX TRANSMISSION
   PROJECT
--------------------------------------------------------- Appendix V:6

The Mead-Phoenix Transmission Project, involving a $94.7 million
investment by Western, including capitalized interest, was intended
to increase the power transmission capability between parts of
Arizona, Nevada, and California.  The project's expected demand has
not materialized, and it is unclear whether Western will be able to
market the project's capacity.  From April 1996, when the project
came into service, through January 1997, it had revenues of only
$71,319, while incurring operation and maintenance and interest
expenses of nearly $7.3 million, resulting in a net loss of about
$7.2 million.  According to Western, if the project does not achieve
the level of sales assumed in the transmission charges, the PMA will
begin a new rate process to ensure recovery of the project's costs. 
Western is considering blending the project's rates into the overall
transmission rates for the Pacific Northwest-Pacific Southwest
Intertie.  If the blending cannot be accomplished, we believe it is
probable that the government will lose at least some of its $94.7
million investment in the Mead-Phoenix project.  Even with the
consolidation, we see no indication that demand for power from the
project will increase or that the PMA will be able to successfully
market the entire transmission capacity, and we therefore conclude
that the risk of future losses to the government is reasonably
possible. 




(See figure in printed edition.)Appendix VI
COMMENTS FROM THE DEPARTMENT OF
ENERGY
=========================================================== Appendix V



(See figure in printed edition.)

See comment 3. 



(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.)



(See figure in printed edition.)


GAO'S COMMENTS

The following are GAO's comments on the letter dated January 28,
1998, from the Department of Energy. 

1.  We do not believe that we have overstated the magnitude of the
net financing costs.  Our methodology for determining such costs is
discussed below under comment 7.  Also, our evaluation of comments
about the recovery of unrecovered costs is discussed below in
comments 22, 23, and 25.  We do not believe that we have overstated
the risk of nonrecovery of some of the federal investment related to
hydropower projects.  Our evaluations of risk levels are discussed
under comments 23, 24, and 25.  We did not add language acknowledging
the use of current interest rates on federal investments since 1983
because that fact was already contained in chapter 2 of our draft
report.  We added language to discuss the PMAs' proposed actions to
recover additional costs to chapters 2 and 3 and appendix V.  We also
added language to chapters 1 and 3 to describe the PMAs' efforts to
reduce costs or otherwise improve their business practices. 

2.  We incorporated changes in the body of the report and to the
executive summary as appropriate. 

3.  We expanded our discussion of the role of public power in chapter
1 to include DOE's views on public power's role in providing
competition for IOUs and in charging power rates against which the
power rates of the IOUs can be compared.  We also revised our
discussion of the mission of federal power in the executive summary
to clarify that rural electrification was not the sole purpose of
selling federal power. 

4.  We added to chapter 1 a description of the PMAs' recent actions
to file tariffs relative to FERC Order 888. 

5.  We added information on actions the PMAs have taken to enhance
their competitiveness, including cost reduction efforts by
Southwestern and Western, to chapter 1.  These actions did not change
the competitiveness of the PMAs enough to warrant changing our
assessments of risk. 

6.  We disagree with DOE's comments on the magnitude of net financing
costs and the degree of risk.  Its comments on net costs are
discussed below in comment 7.  Its comments on the degree of risk are
discussed in comments 23, 24, and 25. 

7.  In commenting on an earlier GAO product, Southeastern,
Southwestern, and Western ("the three PMAs") as well as Bonneville
disagreed with our estimate of the net financing costs.  Two broad
issues were raised:  (1) disagreement with our use of the portfolio
methodology for estimating the net financing costs to the federal
government for appropriated debt, including the use of the weighted
average interest rate on outstanding long-term Treasury bonds, and
(2) the assertion that the PMAs' appropriated debt is analogous to a
mortgage loan.  To calculate the net financing costs to the Treasury
under the portfolio method, we obtained the federal government's
annual interest income from the PMAs by multiplying the amount of the
PMAs' appropriated debt outstanding as of September 30, 1996, by the
weighted average interest rate paid by the PMAs.  To calculate
interest expense for the federal government, we multiplied the amount
of the PMAs' appropriated debt outstanding by the average interest
rate the Treasury was paying on its portfolio of bonds outstanding at
the end of fiscal year 1996--9 percent--which yields an estimate of
the amount of interest expense the Treasury must pay on the PMAs'
outstanding appropriated debt.  The difference between the federal
government's interest income and interest expense represents the net
financing cost.\1 DOE stated that it believes that the use of the
portfolio methodology assumes that both the PMAs' interest rate and
the Treasury's cost of funds are variable, so that the cost
difference on any individual investment varies from year to year.  It
stated that this is equivalent to assuming that the PMAs'
appropriated debt should be refinanced annually.  DOE stated that
comparing the interest rates assigned to PMA financing with the
Treasury's rates in the years the financing was provided
(loan-by-loan methodology) would be a more accurate way of
determining the net financing cost.  DOE and Bonneville also
disagreed about how we estimated the net financing costs on
outstanding appropriated debt by using the interest rate on the
Treasury's outstanding bond portfolio. 

As discussed in GAO/AIMD-97-110A, we define the net financing cost to
the federal government as the difference between the Treasury's
borrowing cost and the interest income received from RUS' borrowers,
the PMAs, and TVA.  Our basic methodology is to determine whether the
federal government received a return sufficient to cover its
borrowing costs and, if not, to estimate the net financing cost. 
RUS, the PMAs, and TVA had several forms of federal debt outstanding
at September 30, 1996.  Each of these forms of federal debt had
different terms and thus required us to apply variations of our basic
methodology in assessing whether a net financing cost existed and, if
so, estimating the amount. 

We continue to believe that, for the PMAs' appropriated debt, the
portfolio methodology best captures the combined impact of the four
distinct aspects of the net financing cost that we identified:  (1)
the difference between the PMAs' borrowing rate and the Treasury's
borrowing rate for securities of similar maturity at the time the
appropriation was made, (2) the PMAs' ability to repay the highest
interest rate debt first, (3) the interest rate risk arising from the
Treasury's general inability to refinance or prepay outstanding debt
in times of falling interest rates, and (4) the difference in the
maturities of the three PMAs' and Bonneville's appropriated debt and
the Treasury's bonds.  We believe that the suggested loan-by-loan
methodology is limited in that it captures only that portion of the
net financing cost arising from the interest rate spread and not the
other three integral aspects of that cost.\2

8.  We revised the executive summary and chapter 2 to state that
maintenance problems differ by location within the operating
agencies. 

9.  We acknowledge that planning and budgeting problems do not
originate solely within the operating agencies; rather, they are
endemic of the federal budgeting process government wide.  However,
we believe that the operating agencies' lengthy and complex processes
contribute to the overall problem.  We clarified chapter 2
accordingly. 

10.  We acknowledge that average revenue per kWh (total
revenues/total electricity sales) is an imperfect indicator of
electricity rates because it combines the costs of several types of
services, such as capacity, peak service, and off-peak service. 
However, for our analysis, it is a strong, broad, indicator of the
relative power production costs of the PMAs compared to IOUs and
publicly owned generators.  For example, TVA's 1997 business outlook
presents the agency's goals for competitive power rates in terms of
overall, systemwide rates.  And, in responding to our August 1995
report,\3 TVA's consultant used this measure in assessing TVA's
competitiveness.  Also, the fiscal year 1995 annual reports for
Southeastern, Southwestern, and Western--our primary data sources for
PMA sales data--reported each customer's total electricity purchases
and revenues.  They do not present the data by type of service.  The
same is true for the non-PMA wholesale data that we received from
EIA.  Moreover, we believe that average revenue per kWh is a proxy
for rates that is widely used in the industry.\4

11.  We added to table 2.1 information reflecting the comments
provided on the risk of project cost recovery, especially the
statements about recovery of costs for units that are not allowed to
operate or to operate as originally designed.  However, we believe
the recovery of costs from nonoperating units overlooks the policy
guidance contained in DOE Order RA 6120.2, which indicates that if
the nonoperational units are not placed into commercial service, the
power customers will not be required to repay the investment. 

12.  We acknowledge that DOE disagrees with some of the findings in
previous GAO products.  We do not agree with its conclusion that our
evaluations are somewhat inaccurate and incomplete.  Our responses to
many of DOE's specific comments on issues raised in past reports are
contained in this appendix.  We have added detailed information
throughout our final report to make our current report more complete. 

13.  We agree that the value of specific federal assets considered
for any divestiture can vary widely based on the assumptions used. 
Our report contains a discussion of the liabilities, assets, and
restrictions that may be retained or transferred by the government
upon any divestiture.  Our report concludes that these factors would
affect the price the government would obtain for its assets. 

14.  We acknowledge that much federal investment is regional in
nature and believe that this condition leaves such investment open to
debate by the Congress and others.  We also acknowledge that an
examination of topics such as why federal power is provided in
certain regions of the nation, transmission limitations, and the
regional equity of all federal government spending could be
undertaken.  Such analysis, however, is outside the scope of this
report. 

15.  We agree that alternative financing does not necessarily reduce
the opportunities for oversight by the Congress and others.  We added
language to chapter 3 describing congressional latitude in fostering
opportunities for oversight under expanded use of alternative
financing. 

16.  We disagree with DOE's conclusion that we present the option of
charging rates based on competition only because of the electric
utility industry's trend toward competitive pricing.  We agree that
many consumers expect lower prices as a result of a restructured
electrical industry and that competitive pricing will most likely
lead to higher prices for most PMA customers.  After the wholesale
market restructures, competitive rates may still be higher than the
rates the PMAs currently charge. 

17.  We expanded our discussion of cost reallocation in chapter 3 to
recognize that an equity issue exists concerning power purchasers
having to repay costs that are not related to power. 

18.  We revised chapter 3 to state that Western does not support
corporatization of that marketing program at this time. 

19.  We revised the executive summary and chapter 3 to include the
comment that the degree of oversight following any corporatization
depends on the particular arrangements chosen by the Congress for
itself or outside oversight agencies. 

20.  We added to chapter 3 additional language concerning Native
American interests in rights-of-way based on this review. 

21.  We agree that divestitures would include federal transaction
costs.  If we are requested to analyze the costs of quantifying the
benefits and costs associated with divesting the federal hydropower
assets, we would consider quantifying the transaction costs. 

22.  DOE stated that some portion of the $518 million will be
recovered even if the pumping units are never commercially operated. 
We added to table 2.1 and appendix V DOE's assertion that some
unspecified portion of the $518 million investment in pumping units
at the Russell project will be recovered even if the units are never
commercially operated.  However, we added language that we believe
this assertion by the PMAs overlooks the policy guidance contained in
DOE Order RA 6120.2, which indicates that if the nonoperational units
are not placed into commercial service, the power customers will not
be required to repay the investment. 

23.  We added to table 2.1 and appendix V DOE's assertion that
Southwestern can add to its power repayment study the power-related
costs of pumpback units at the Truman project even if the units are
never operable.  We also added, however, that we do not believe that
a change in risk category is appropriate until these costs are
actually added to Southwestern's repayment study. 

24.  We added to chapter 2 and appendix V language describing the
January 1998 decision by FERC that approved the lowering of rates for
power marketed from the Central Valley Project (CVP).  We disagree
that this action is sufficient to warrant an upgrading of the risk
category to "remote."

25.  We reviewed the comments provided for an earlier GAO report,
GAO/AIMD-97-110A, and believe that our earlier evaluations are
accurate.  For example, we continue to (1) agree that CVP was able to
meet its repayment obligations in fiscal year 1996, (2) believe that
the Central Valley Project Improvement Act may adversely affect the
availability of water for power generation,\5 and (3) conclude that
the appropriate category of risk for CVP is "reasonably possible."
Assignment of this risk category is the result, in part, of the
uncertain potential reductions in the Trinity River's water flows
available to the CVP, as DOE noted.\6

26.  We disagree that the risk of nonrepayment for the Washoe Project
is remote.  DOE states that Western staff are proposing the blending
of the costs of power from the Washoe Project with the costs of power
from the CVP after 2004.  This proposal was noted in appendix V of
our draft report.  We continue to believe that the risk of
nonrecovery is probable, if this proposal is not implemented, and
that the risk category improves only to reasonably possible, if the
proposal is implemented.  We believe that the risk associated with a
blended rate is not remote because, as we state in appendix V, the
CVP itself faces challenges in remaining competitive. 



(See figure in printed edition.)Appendix VII

--------------------
\1 For a further discussion of PMA financing, see GAO/AIMD-96-145. 

\2 A more complete discussion of our methodology is contained in
GAO/AIMD-97-110. 

\3 GAO/AIMD/RCED-95-134. 

\4 More discussion of our use of average revenue per kWh is contained
in GAO/AIMD-97-110. 

\5 For example, according to the Bureau's comments on the draft
report, managing 800,000 acre-feet of water within the CVP to benefit
the environment could reduce hydropower generation by 5 percent. 

\6 A more detailed discussion of the risk of nonrecovery of the
federal investment associated with the CVP is contained in
GAO/AIMD-97-110A. 


COMMENTS FROM THE DEPARTMENT OF
THE INTERIOR
=========================================================== Appendix V



(See figure in printed edition.)



(See figure in printed edition.)


GAO'S COMMENTS

The following are GAO's comments on the Department of the Interior's
letter dated February 3, 1998. 

The Department of the Interior (Interior) provided us with comments
that were intended to clarify the respective roles and relationship
between the Bureau of Reclamation and the PMAs and to clarify
statutory requirements for the management of the Bureau's facilities
for multiple purposes. 

1.  In its letter, Interior stated its concern that our report
concluded that the availability factor for the Bureau's power plants
is lower than an average for an industry benchmarking group because
the Bureau has not adequately maintained and repaired its power
plants.  In response to this comment, we revised the report,
including the executive summary, to state that the federal planning
and budgeting processes, as implemented by the Bureau, do not allow
for timely funding of needed repairs to the Bureau's power plants. 
This situation, in turn, has contributed to the decreased
availability of the Bureau's power plants.  We believe this revision
is supported by the Bureau's own performance data for its power
plants and by the fact that the Bureau is negotiating and has
negotiated arrangements with Western and Western's preference
customers for those customers to provide advanced funding of needed
repairs.  This arrangement would allow funding to occur when needed
to pay for repairs to the power assets. 

In addition, while acknowledging the importance of irrigation and
other multiple purposes as key factors in managing the Bureau's water
and power resources, we disagree that the need to balance multiple
purposes would necessarily decrease the availability of the Bureau's
power plants to generate power.  In this regard, the availability
factor does not measure how much water is diverted for multiple
purposes.  This factor simply expresses the amount of time a plant is
available to generate power divided by the total number of hours in a
time period.  In addition, many other federal and nonfederal power
plants also generate power subject to multiple uses of the water. 
Yet the Bureau's availability factor was below that of this
comparison group. 

Interior also provided us with general and detailed comments.  Our
responses to the general comments are contained below.  The detailed
comments were of a technical or editorial nature, which we addressed
as appropriate in the report. 

2.  The report currently discusses alternative financing arrangements
in the Bureau's Central Valley Project (CVP) and Loveland projects. 
We expanded our discussion of these arrangements in the executive
summary, chapter 2, and chapter 3. 

3.  We disagree with Interior's statement that we characterize
insufficient resources to fund repairs as a sign of "management
deficiency." We do not characterize the Bureau's resource levels in
this fashion.  Rather, we state that the budget process used by the
Bureau, the Corps, and other federal agencies is not always
appropriate for a commercial activity.  In the case of the Bureau and
the Corps, it does not deliver funding on a predictable, timely basis
when it is needed to pay for repairs to the federal hydropower
assets.  We modified the report to emphasize that the planning and
budgeting processes are causes for the relatively low availability
factor of the Bureau's power plants. 

We also disagree with Interior's statement that the Bureau has been
able to adequately manage the appropriations to ensure that necessary
repairs are made.  The Bureau's staff at the regional level,
Western's staff, and Western's preference customers contend that
obtaining funding for necessary repairs to the Bureau's power assets
is sometimes difficult and unpredictable, amid shrinking budgets. 
For example, the General Manager of the Northern California Power
Agency, on March 19, 1996, testified before the House Subcommittee on
Water and Power that: 

     ".  .  .  OM&R [operation, maintenance, and rehabilitations]
     problems have occurred in the CVP and, assuming the
     discretionary spending portion of the federal budget continues
     to shrink as agreed by the Congress and by the Executive Branch,
     we will get worse with age.  However, in the case of the CVP's
     Shasta Dam, a creative customer-financing solution has been
     implemented to the aging problems with three of the five
     generators in the dam."

In another example, a study conducted by Northern California Power
Agency, the Sacramento Municipal Utility District, the Bureau, and
Western in September 1996 details shortfalls within the CVP in the
maintenance and operating condition of the CVP's power plants.  The
report, based on walk-through inspections of the Bureau's power
plants by teams of engineers from the Bureau, Western, and the
Northern California Power Agency, states in its executive summary,
among other things,

     "The majority of the original CVP facilities that are operated
     and maintained by Reclamation [the Bureau] were constructed over
     a period from the mid 1940s to the late 1970s.  A significant
     amount of plant equipment is obsolete and replacement parts are
     no longer available.  Other equipment is one-of-a-kind type of
     hardware or built by vendors who have moved into new technology
     or are no longer in business.  Much of the original equipment
     and systems are well worn and require an abnormal level of
     maintenance to keep the facilities in operation.  Several
     generator units have been or are being upgraded, but the
     majority of the units are long into their life cycle and soon
     will need attention to continue at rated operation."

In addition, according to the Northern California Power Agency's
General Manager and officials from Western's and other PMAs' customer
organizations, shrinking budget levels and the unpredictability of
funding levels have led to alternative funding methodologies whereby
PMA customers donate funds to pay for needed repairs, upgrades, and
rehabilitations.  In our view, such measures are becoming
increasingly popular among the operating agencies, the PMAs, and
their customers in order to ensure that the federal power resource is
adequately maintained and repaired and the PMAs' preference customers
receive the power in a manner to which they are accustomed. 

4.  We believe that the Bureau operates and maintains its power
plants within the constraints posed by its budget and is trying to do
so in a more efficient, businesslike manner.  We further agree, as
previously discussed with Bureau officials, that the Bureau has no
formal policy of deferring maintenance of its power assets.  In this
regard, we revised the report where appropriate. 

As stated before, we disagree that the need to balance multiple
purposes would necessarily decrease the availability of the Bureau's
power plants to generate power.  In this regard, the availability
factor does not measure how much water is diverted for multiple
purposes.  This factor simply expresses the amount of time a plant is
available to generate power divided by the total number of hours in a
time period.  In addition, many other federal and nonfederal power
plants also generate power subject to multiple uses of the water. 
Yet the Bureau's availability factor was below that of this
comparison group. 

5.  Interior provided information on interagency arrangements to fund
O&M for federal power plants in the Pacific Northwest.  We revised
our report as suggested. 

6.  According to Interior, the executive summary implies that all
power generated by the Bureau is marketed by the PMAs, ignoring the
fact that power generation is a "secondary purpose" and the Bureau
uses 5 percent to 10 percent of the power for project purposes.  The
PMAs market the remaining power.  We agree with these statements and
added language to the executive summary and chapter 1 to recognize
that the PMAs market only the power that remains after it has been
used for project purposes--for example, to pump water for irrigation. 

7.  According to Interior, our report prescribes a "businesslike"
approach to the PMAs' rate-setting practices that would maximize
revenues to recover investments, much like IOUs set rates.  Interior
adds that the rate-setting practices of the PMAs should be examined
in light of the multiple purposes served by the Bureau's water
projects.  Interior stated that power generation is a byproduct of
the federal irrigation projects.  It also stated that power
generation, along with power revenues, is maximized subject to the
multiple purposes of the projects.  Moreover, power revenues are
intended to pay for the features of the projects, according to
Interior. 

We disagree with Interior's comments that the report prescribes an
approach to the PMAs' rate-setting practices that would maximize
revenues to recover investment.  In fact, the report makes no
recommendations that can be construed as "prescribing" any one
approach.  Rather, it presents options that policymakers may consider
to better capture the federal investment in power-related facilities
as well as those federal investments allocated to power for
repayment.  In describing these options, we took great care to ensure
that we reflected many of the options' pros and cons.  For example,
we state that the PMAs could opt to increase their power rates to
repay the federal investment faster.  However, to counterbalance that
advantage, we carefully state that any movement by the PMAs to
increase their rates could make the PMAs' power over-priced in
evolving competitive markets.  Overpriced power would be difficult to
sell, thus jeopardizing the repayment of the federal investment.  In
addition, the report recognizes that power sold by the PMAs is
generated and marketed subject to the multiple purposes of water
projects.  The report also states that power is used for project
purposes and also recognizes that power revenues pay for nonpower
features--for instance, we discuss the concept of aid-to-irrigation
and that power revenues are scheduled to repay about 70 percent of
the capital costs associated with irrigation projects.  However, on
the basis of the Bureau's comments, we revised the executive summary
and chapter 1 to emphasize the first use of power for project
purposes--for example, irrigation. 

Interior also stated that our report characterizes the Bureau as
operating in an unsound fashion because all costs are not recovered
through the PMAs' rates.  We disagree with this statement.  Nowhere
does the report state that the Bureau operates its projects in an
unsound fashion.  In fact, the report explicitly states that the PMAs
are following applicable laws and regulations in setting rates and
recovering costs.  For example, in our discussion of cost recovery
within the Pick-Sloan program, chapter 2 and appendix V clearly state
that suballocated costs\1 will not be recovered absent congressional
action because the current repayment methodology and suballocation
amounts are based on federal statutes.  In connection with the Shasta
Project, the executive summary clearly recognizes that "the 1991
Energy and Water Development Appropriations Act specified that these
costs not be allocated to power for repayment through PMA customers'
electric rates." This point is made in chapter 2, also.  The report
does not imply in either the Pick-Sloan or the Shasta case that the
agencies are intentionally deferring cost recovery through power
rates. 

8.  We disagree with Interior's comment that the options in our draft
report recommended diverging from the fundamental policy that
encourages or requires the separation of various electric utility
aspects and services.  First, our draft report contained no
recommendations.  Second, our draft report clearly stated, "Although
the electric utility industry is now unbundling its services,
depending on how a government corporation was structured, the
generation, transmission, and marketing aspects could be put under
one agency, possibly reducing overhead." Therefore, we made no change
to the final report. 

9.  We agree that the Bureau and the Corps have separate
organizations, management, missions, standards, policies, laws, and
regulations.  However, no changes are needed to the report.  We only
refer to the Corps and the Bureau together primarily when addressing
their common role as operating agencies.  Where appropriate, for
example, when addressing the availability factors of the Bureau's and
the Corps' power plants, we differentiate between the agencies. 



(See figure in printed edition.)Appendix VIII

--------------------
\1 For a further discussion of suballocated costs, see
GAO/T-RCED-96-142. 


COMMENTS FROM THE DEPARTMENT OF
DEFENSE
=========================================================== Appendix V



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


GAO'S COMMENTS

The following are GAO's comments on the U.S.  Army Corps of
Engineer's letter dated January 29, 1998.  The Corps provided GAO
with detailed, technical, and editorial comments in response to our
report. 

1.  In connection with the executive summary, the Corps noted that
helping electrify rural America was only one of the purposes of
selling federal power and that cost recovery was the primary purpose
of generating hydropower.  The Corps added that hydropower generation
is generally a secondary purpose in multipurpose federal water
resource projects; flood control, navigation, and irrigation are the
primary purposes.  In response, we clarified the executive summary to
better reflect that selling federal power in rural areas is only one
of several missions of the federal power program.  The executive
summary already stated that other purposes exist, such as flood
control, navigation, and irrigation.  We also revised the executive
summary to state that hydropower sold by the PMAs is that which
remains after it has been used for project purposes, such as pumping
water to the fields being irrigated. 

2.  The Corps noted that prior to fiscal year 1995, the pension and
postretirement benefits of power-related federal employees were not
made available to the federal power agencies.  However, when the
Office of Personnel Management stopped budgeting for these costs, the
federal agencies became responsible for them, according to the Corps. 
The Corps states that the report implies that the federal power
agencies were "knowingly avoiding these costs." We disagree with the
Corps' assessment and therefore made no revisions to the report.  The
report clearly states that the federal power agencies are recovering
costs "under current federal laws, an applicable DOE order, and
repayment practices," and it also notes that the PMAs were generally
following applicable laws and regulations in their rate-setting
practices. 

3.  We incorporated the editorial revision suggested. 

4.  On the basis of updated information provided by the Corps, we
revised table 2.1 and appendix V to update the status of repairs made
at the Russell and Truman projects. 

5.  The Corps stated that we did not use more recent data reflecting
the improved performance of its power plants.  We revised the
executive summary and chapter 2 to include this new information. 

6.  The Corps stated that the report did not mention its major
rehabilitation program, which dedicated funding of $450 million
through fiscal year 2007 to repair the Corps' power plants.  A Corps
official attributed part of the improved availability of the Corps'
hydropower plants to this program.  We revised chapter 2 to include
this new information. 

7.  The Corps remarked that none of our options would necessarily
reduce oversight by the Congress or by the administration.  We agree
and revised the executive summary and chapter 3 to state that the
extent of external oversight would depend on how the options are
structured and added that this oversight could be provided by the
Congress or by the Office of Management and Budget by requiring the
power agencies to submit expenditure data. 

8.  The Corps stated that power is rarely marketed and priced on the
basis of one project, but is marketed and priced on the basis of a
system.  In response, we revised the executive summary to indicate
that power is marketed and sold from rate-setting systems. 

More importantly, however, the Corps added that even if a project
does not "pull its weight," the system overall will continue to be
economical.  However, we believe that a high cost generating project
within a rate-setting system, when combined with such factors as the
need to mitigate environmental impacts, can cause rates to increase
to levels that equal or even exceed regional market rates for
wholesale power.  If power rates become uncompetitive, the
government's ability to sell its power, and hence to repay its
investment, is diminished.  For example, the composite rates of the
Colorado River Storage Project and Central Valley Project have
experienced upward pressures, in part as a result of the need to
mitigate environmental impacts, to the point that these projects'
rates are approaching regional rates for certain types of power. 

9.  We made the editorial revision suggested. 

10.  The Corps suggests that information should be included in the
report as to why the issue of recovering the annual costs of pension
and postretirement health benefits came to light.  In September 1996,
a GAO report identified costs incurred by the federal government to
generate, transmit, and market power.\1 The issue the Corps referred
to came to light because this report found that these costs were not
being recovered through the PMAs' rates.  Therefore, we did not
revise our report. 

11.  The Corps states that certain costs associated with the Truman
project cannot be recovered pending the lifting of a court
injunction.  The Corps suggests that we revise the report to state
that a commercial power producer, in the same position, would also
not be allowed to include these costs in its power rate case.  We did
not revise the report because the ability of an IOU to ultimately
include those costs in its rates would depend on the actions of a
public utility commission, which would be uncertain. 

12.  The Corps provided new information about its $450 million major
rehabilitation program that we incorporated into chapters 2 and 3. 

13.  In chapter 3, we included information provided by the Corps that
the Army's General Counsel has determined that the Corps can accept
funds from power customers, with certain restrictions. 

14.  The Corps stated that the memorandum from the Army's General
Counsel stated that it would be desirable to have specific
legislation clarify the authorities but that the law allows certain
contributions.  We did not revise the report because the text already
contained this information. 

15.  The Corps stated that the report, in its discussion of
divestiture, does not discuss the transfer of federal liabilities to
new owners.  We did not revise the report because it already
discussed in depth the multipurpose aspects of water projects and the
impact on a divestiture of the need to manage water for these
purposes. 

16.  We incorporated the editorial revision suggested. 

17.  The Corps stated that if the powerhouses are sold and the
government continues to operate the balance of a water project, the
taxpayers' future liabilities would be great.  We did not revise our
report in response to this observation because the report already
addressed the trade-offs that would have to be considered by
policymakers as they decide whether and how to proceed in a
divestiture of the federal hydropower assets. 

18.  In connection with our discussion of potential rate increases
after a divestiture, the Corps stated that the report engages in
"energy pricing." The Corps said that as electric services are
unbundled, the generating capacity may be more valuable than the
electric energy, with an additional impact on the PMAs' customers. 
We decline to revise our report because it already recognized in a
footnote in chapter 3 that some PMA customers already use the PMAs'
power primarily to satisfy demand during peak periods.  For these
customers, in the event of a divestiture, the impacts on their rates
may be higher than if they had not relied primarily on the PMAs'
power to serve their demand during peak periods. 

19.  The Corps provided us with new information on the nameplate
capacity of the Truman project and the status of repairs on the
project's generating units that we incorporated in table 2.1 and
appendix V. 



(See figure in printed edition.)Appendix IX

--------------------
\1 Power Marketing Administrations:  Cost Recovery, Financing, and
Comparison to Nonfederal Utilities (GAO/AIMD-96-145, Sept.  19,
1996). 


COMMENTS FROM THE BONNEVILLE POWER
ADMINISTRATION
=========================================================== Appendix V


GAO'S COMMENTS

The following are GAO's comments on the Bonneville Power
Administration's letter dated January 27, 1998. 

Bonneville repeated several points it had presented in comments on
our September 1997 report.\1 Specifically, Bonneville (1) disagreed
with our position that its operations entail substantial net costs,
(2) reiterated that it continues to believe that satisfying its
current repayment obligations on balance will provide full
compensation for the appropriated investments of the Columbia River
power system, (3) contended that our position on net costs does not
use a true measure of the interest cost to the government, (4) stated
that we ignore recent legislation that confirms the Congress' belief
as to the adequacy of Bonneville's repayment responsibilities, and
(5) asserted that we underplay the significant financial implications
of the public benefits funded by Bonneville.  Our position on these
issues is unchanged. 

In connection with the first point about net costs, we found in our
September 1997 report that Bonneville had incurred substantial debt
at below-Treasury interest rates, as shown in several examples in the
subject report.  We also noted that Bonneville is only required to
pay outstanding principal on the year of maturity and that Bonneville
is allowed to repay appropriated debt with the highest interest rate
first and to keep the appropriated low-interest rate on its books for
decades.  We also point out that in fiscal year 1996, the Treasury
incurred a net financing cost of $377 million as a result of
Bonneville's activities.  This net negative cash flow to the federal
government will continue as long as the appropriated debt and
corresponding Treasury debt are outstanding. 

We continue to disagree with Bonneville's second point about its
current repayment obligations providing full compensation for the
appropriated investments of the Columbia River power system.  As
discussed above, not only do Bonneville's operations entail a net
financing cost, but they also incurred net costs related to
postretirement benefits for its employees.  Our report already
acknowledged in chapter 3 that Bonneville plans to begin recovering
these costs in fiscal year 1998, with full recovery planned beginning
in fiscal year 2002.  Consistent with current policies and law, the
PMAs do not plan to recover pre-fiscal year 1998 net costs. 
Bonneville in its comments provided no new information that would
cause our position to be changed. 

We also continue to disagree with Bonneville's third point that our
position on net costs does not use a true measure of the interest
cost to the government.  Among other points we made in replying to
Bonneville's comments on our September 1997 report, the interest rate
that Bonneville is to pay on its appropriated debt under the Omnibus
Consolidated Rescissions and Appropriations Act of 1996 supports our
position that a long-term Treasury rate is the correct rate to use in
our portfolio analysis.  Under the act, that interest rate is based
on long-term Treasury interest rates. 

Bonneville's fourth point is that we ignore recent legislation that
confirms the Congress' belief as to the adequacy of Bonneville's
repayment responsibilities.  We have no way of ascertaining the
Congress' beliefs about the adequacy of Bonneville's repayment
responsibilities.  Moreover, reporting on, evaluating, or commenting
on the congressional view was beyond the scope of this review. 
Therefore, we declined to revise the report. 

Bonneville's fifth point is that we underplay the significant
financial implications of the public benefits funded by Bonneville. 
We decline to revise the report in response.  The scope of this
assignment did not include examining the public benefits that
Bonneville and the other PMAs provide to their respective regions. 
However, it should be noted that the report states that water
projects entail a number of multiple purposes, and hence benefits to
the public, such as providing for navigation, flood control, and
irrigation.  The report also notes that in the event of a
divestiture, these purposes may continue as federal functions--a
factor that would have to be considered by policymakers in deciding
if and how to divest the federal hydropower assets. 



(See figure in printed edition.)Appendix X

--------------------
\1 GAO/AIMD-97-110 and 110A. 


COMMENTS FROM THE FEDERAL ENERGY
REGULATORY COMMISSION
=========================================================== Appendix V



   GAO'S COMMENTS
--------------------------------------------------------- Appendix V:7

The following are GAO's comments on FERC's letter dated January 26,
1998. 

FERC provided us with comments on how it would regulate the federal
hydropower assets after a divestiture and the impact on available
generating capacity as a result of relicensing nonfederal hydropower
plants for which applications were filed in 1991.  FERC stated that
the position of its staff was that FERC did not want to license any
divested federal hydropower assets on a basis that excludes some of
the project's features that have a role in power production.  FERC
also stated that it would be able to regulate any divested assets
because it had experience regulating the multipurpose aspects of over
1,600 nonfederal hydropower plants.  FERC added that in its
relicensing of 157 applications filed in 1991, the projects that were
relicensed experienced a slight increase in total capacity available
to generate power but a slight decline in actual generation.  Our
report was revised to address these and other suggestions. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================== Appendix XI

RESOURCES, COMMUNITY, AND ECONOMIC
DEVELOPMENT DIVISION

Philip Amon
Ernie Hazera
Charles Hessler
Susan Kladiva
Peg Reese
Daren Sweeney
Martha Vawter

OFFICE OF THE GENERAL COUNSEL

Jackie Goff





RELATED GAO PRODUCTS
============================================================ Chapter 1

Deferred Maintenance Reporting:  Challenges to Implementation
(GAO/AIMD-98-42, Jan.  30, 1998). 

Rural Utilities Service:  Opportunities to Operate Electricity and
Telecommunications Loan Programs More Effectively (GAO/RCED-98-42,
Jan.  21, 1998). 

Federal Electricity Activities:  The Federal Government's Net Cost
and Potential for Future Losses (GAO/AIMD-97-110 and 110A, Sept.  19,
1997). 

Deferred Maintenance:  Reporting Requirements and Identified Issues
(GAO/AIMD-97-103R, May 23, 1997). 

Bureau of Reclamation:  Reclamation Law and the Allocation of
Construction Costs for Federal Water Projects (GAO/T-RCED-97-150, May
6, 1997). 

Rural Development:  Financial Condition of the Rural Utilities
Service's Loan Portfolio (GAO/RCED-97-82, Apr.  11, 1997). 

Federal Power:  Issues Related to the Divestiture of Federal
Hydropower Resources (GAO/RCED-97-48, Mar.  31, 1997). 

Privatization:  Lessons Learned by State and Local Governments
(GAO/GGD-97-48, Mar.  14, 1997). 

Budget Issues:  Budgeting for Federal Capital (GAO/AIMD-97-5, Nov. 
12, 1996). 

Power Marketing Administrations:  Cost Recovery, Financing, and
Comparison to Nonfederal Utilities (GAO/AIMD-96-145, Sept.  19,
1996). 

Northwest Power Planning Council:  Greater Public Oversight of
Business Operations Would Enhance Accountability (GAO/RCED-96-226,
Aug.  30, 1996). 

Federal Power:  Outages Reduce the Reliability of Hydroelectric Power
Plants in the Southeast (GAO/T-RCED-96-180, July 25, 1996). 

Federal Power:  Recovery of Federal Investment in Hydropower
Facilities in the Pick-Sloan Program (GAO/T-RCED-96-142, May 2,
1996). 

Budget Issues:  Privatization/Divestiture Practices of Other Nations
(GAO/AIMD-96-23, Dec.  15, 1995). 

Government Corporations:  Profiles of Existing Government
Corporations (GAO/GGD-96-14, Dec.  13, 1995). 

Federal Electric Power:  Operating and Financial Status of DOE's
Power Marketing Administrations (GAO/RCED/AIMD-96-9FS, Oct.  13,
1995). 

Tennessee Valley Authority:  Financial Problems Raise Questions About
Long-term Viability (GAO/AIMD/RCED-95-134, Aug.  17, 1995). 

Government Corporations:  Profiles of Recent Proposals
(GAO/GGD-95-57FS, Mar.  30, 1995). 

Naval Petroleum Reserve:  Opportunities Exist to Enhance Its Value to
the Taxpayer (GAO/T-RCED-95-136, Mar.  22, 1995)

Uranium Enrichment:  Observations on the Privatization of the United
States Enrichment Corporation (GAO/T-RCED-95-116, Feb.  24, 1995). 

Deficit Reduction:  Experiences of Other Nations (GAO/AIMD-95-30,
Dec.  13, 1994). 

Uranium Enrichment:  Activities Leading to Establishment of the U.S. 
Enrichment Corporation (GAO/RCED-94-227FS, June 27, 1994). 

Bonneville Power Administration:  Borrowing Practices and Financial
Condition (GAO/AIMD-94-67BR, Apr.  19, 1994). 

Budget Issues:  Incorporating an Investment Component in the Federal
Budget (GAO/AIMD-94-40, Nov.  9, 1993). 

Electricity Regulation:  Electric Consumers Protection Act's Effects
on Licensing Hydroelectric Dams (GAO/RCED-92-246, Sept.  18, 1992). 

Federal Electric Power:  Views on the Sale of Alaska Power
Administration Hydropower Assets (GAO/RCED-90-93, Feb.  22, 1990). 

Lessons Learned About Evaluation of Federal Asset Sale Proposals
(GAO/T-RCED-89-70, Sept.  26, 1989). 

Policies Governing Bonneville Power Administration's Repayment of
Federal Investment Still Need Revision (GAO/RCED-84-25, Oct.  26,
1983). 

Triennial Assessment Of The Tennessee Valley Authority--Fiscal Years
1980-1982 (GAO/RCED-83-123, Apr.  15, 1983). 

Congress Should Consider Revising Basic Corporate Control Laws
(GAO/PAD-83-3, Apr.  6, 1983). 

Tennessee Valley Authority--Options For Oversight (GAO/EMD-82-54,
Mar.  19, 1982). 


*** End of document. ***