Capital Financing: Potential Benefits of Capital Acquisition	 
Funds Can Be Achieved through Simpler Means (08-APR-05, 	 
GAO-05-249).							 
                                                                 
CAFs have been discussed as a new mechanism for financing federal
capital assets. As envisioned, CAFs would have two goals. First, 
CAFs would potentially improve decision making by reflecting the 
annual cost for the use of capital in program budgets. Second,	 
they would help ameliorate at the subunit level the effect of	 
large increases in budget authority for capital projects (i.e.,  
spikes), without forfeiting congressional controls requiring the 
full cost of capital assets to be provided up-front. Through	 
discussions with budget experts and by working with two case	 
studies, the Departments of Agriculture and of the Interior, we  
are able to describe in this report (1) how CAFs would likely	 
operate, (2) the potential benefits and difficulties of CAFs,	 
including alternative mechanisms for obtaining the benefits, and 
(3) several issues to weigh when considering implementation of	 
CAFs.								 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-05-249 					        
    ACCNO:   A21137						        
  TITLE:     Capital Financing: Potential Benefits of Capital	      
Acquisition Funds Can Be Achieved through Simpler Means 	 
     DATE:   04/08/2005 
  SUBJECT:   Assets						 
	     Budget administration				 
	     Budget authority					 
	     Budget controllability				 
	     Capital						 
	     Cost analysis					 
	     Federal funds					 
	     Financial management				 
	     Funds management					 
	     Internal controls					 
	     Strategic planning 				 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-05-249

United States Government Accountability Office

GAO	Report to the Chairman, Subcommittee on Government Management, Finance, and
    Accountability, Committee on Government Reform, House of Representatives

April 2005

CAPITAL FINANCING

Potential Benefits of Capital Acquisition Funds Can Be Achieved through Simpler
                                     Means

                                       a

GAO-05-249

[IMG]

April 2005

CAPITAL FINANCING

Potential Benefits of Capital Acquisition Funds Can Be Achieved through Simpler
Means

                                 What GAO Found

Capital acquisition funds (CAF) have been suggested as department-level
funds that would use appropriated up-front borrowing authority to buy new
departmental subunit assets. These subunits would then pay the CAF a
mortgage payment sufficient to cover the principal and interest payment on
the Treasury loan. The CAF would use those receipts only to repay Treasury
and not to finance new assets. If existing capital assets were transferred
to the CAF, subunits would pay an annual capital usage charge to the CAF.

CAFs might achieve the goals intended, but these goals can be achieved
through simpler means. Alternative mechanisms, such as asset management
systems, cost accounting systems, and working capital funds may achieve
the goal of allocating annual capital costs and improving decision making
for capital assets. Our case study agencies generally did not indicate
problems with budget authority spikes. They budget in useful segments, use
accumulated no-year authority, or finance capital assets using working
capital funds. Many concerns about CAFs were raised, including the
longterm feasibility of making fixed annual mortgage payments and the
added complexity CAFs would create.

Implementation would raise a number of issues. If CAFs were applied only
to new assets going forward, all programs would not reflect the full
annual cost of capital for decades. Yet the difficulties of including
existing capital are numerous. Even if these issues were tackled, there is
little assurance that CAFs alone would create new incentives for programs
to reassess their use of capital since CAF payments would not affect the
deficit.

Implementation issues could overwhelm the potential benefits of a CAF.
More importantly, current efforts under way in agencies would reflect
asset costs as part of program costs without introducing the difficulties
of a CAF. As long as alternative efforts uphold the principle of up-front
funding, CAFs do not seem to be worth the implementation challenges they
would create. Except for OMB, agencies generally agreed with our
conclusions.

     Up-Front Financing of Federal Capital Assets under a CAF United States
                        Government Accountability Office

Contents

  Letter

Results in Brief
Background
Scope and Methodology
CAF Operations Would Create a New Financing System and New

Oversight Responsibilities CAF Benefits Can Be Achieved through
Alternative Means Without the Added Budget Complexity Several Issues to
Weigh When Considering Implementation

of CAFs Conclusion Agency Comments and Our Response

                                                                     1 2 8 10

12

17

32 40 41

Appendixes                                                              
               Appendix I:  Comments from the Department of the Treasury   44 
              Appendix II:  Comments from the Department of the Interior   47 
                           Figure 1: Up-Front Financing of Federal Capital    
    Figures                Assets before and after Establishing CAFs: New  
                                      Asset Obtained in Year 1              3
                                 Figure 2: NPS Asset Management Plan       19 
                             Figure 3: Illustration of Budget Spikes and   
                                         Potential Smoothing               
                                       Effects of a CAF at ARS             24 

Contents

Abbreviations

ABC activity-based costing
APHIS Animal and Plant Health Inspection Service
ARS Agricultural Research Service
BA budget authority
BLM Bureau of Land Management
BPA Bonneville Power Administration
CAF capital acquisition fund
CBO Congressional Budget Office
CPAIS Corporate Property Automated Information System
DOD U.S. Department of Defense
DOI U.S. Department of the Interior
FASAB Federal Accounting Standards Advisory Board
FBF Federal Buildings Fund
FMS Financial Management Service
FS Forest Service
FTS Federal Technology Service
GSA General Services Administration
INFRA Infrastructure
IT information technology
NASA National Aeronautics and Space Administration
NPS National Park Service
NRC National Research Council
OMB Office of Management and Budget
PBS Public Buildings Service
USDA U.S. Department of Agriculture
WCF working capital fund

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

A

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

April 8, 2005

The Honorable Todd Russell Platts

Chairman

Subcommittee on Government Management, Finance, and Accountability
Committee on Government Reform House of Representatives

Dear Mr. Chairman:

This report responds to your request that we explore the concept of
capital acquisition funds (CAF) as a possible way to reflect the annual
cost for the use of capital in program budgets while still maintaining
up-front congressional control over budgetary resources. To fully
understand how these funds would work in the federal government, we agreed
to (1) describe how CAFs might work as a financing approach for federal
agencies' capital investment, (2) examine the potential benefits and
difficulties of instituting and using CAFs, and (3) identify any issues
that Congress might consider before instituting CAFs.

Since CAFs do not currently exist, we developed an in-depth understanding
of the CAF mechanism through a review of written CAF proposals and
interviews with budget experts at the Office of Management and Budget
(OMB) and the Congressional Budget Office (CBO). We then presented a
detailed description of CAFs to officials at the Departments of
Agriculture (USDA) and of the Interior (DOI) and some of their agencies.
We obtained their insights and opinions on the applicability of the
mechanism at their agencies and on how CAFs might compare to their current
practices for financing capital assets. We also spoke with Department of
the Treasury (Treasury) and General Services Administration (GSA)
officials and congressional staff to gain their perspective on CAFs. Our
work was conducted in Washington, D.C., from May 2004 through January 2005
in accordance with generally accepted government auditing standards. We
obtained comments on a draft of this report from OMB, Treasury, GSA, DOI,
and USDA. Written comments from Treasury and DOI are reproduced in
appendix I and II. We have incorporated technical comments as appropriate
throughout the report.

Results in Brief	In recent years, various interested parties1 have
reported on the different ways that federal agencies can improve planning,
budgeting, and decision making for capital assets. One common
recommendation has been to consider the creation of new capital asset
financing mechanisms, CAFs. The theory behind this recommendation is that
CAFs would both help improve decision making by allocating capital costs
to subunits2 of departments on an annual basis and also help alleviate, at
the subunit level, the large, year-to-year increases in budget authority
(BA)3 (referred to in this report as "spikes") that sometimes occur with
up-front financing of capital assets without changing the requirement for
up-front funding at the department level.

CAFs would operate at the department level and would to some extent
complicate the current process for financing federal capital assets.
Instead of providing departmental subunits with up-front appropriations to
directly purchase capital assets, Congress would appropriate up-front
borrowing authority to the departmental CAF for the full cost of an
asset.4 The CAF then would borrow from the Treasury's general fund to
acquire the asset. The subunit would receive the asset and make a
"mortgage payment"5 to the CAF. This mortgage payment would then be
forwarded to Treasury as repayment for the borrowed funds; it would not be
used to finance new assets. If existing capital assets were transferred to
the CAF, the CAF

1These interested parties include the President's Commission on Capital
Budgeting, OMB, CBO, GAO, and others referenced throughout this report.

2Throughout this report we use the term "subunit" to mean any agency,
bureau, or program that falls under the jurisdiction of a department-level
entity. For example, the Agricultural Research Service would be a subunit
within USDA. We sometimes use the word agency in place of the term
subunit.

3Budget authority is authority provided by law to enter into financial
obligations that will result in immediate or future outlays involving
federal government funds. An appropriation act is the most common means of
providing budget authority. The basic forms of budget authority include
(1) appropriations, (2) borrowing authority, (3) contract authority, and
(4) authority to obligate and expend offsetting receipts and collections.

4Authority to borrow is a type of budget authority and thus is subject to
congressional control. It refers to the statutory authority that permits a
federal agency to incur obligations and make payments to liquidate
obligations out of borrowed monies. This does not include the Treasury's
authority to borrow from the public or other sources under 31 U.S.C. 31.
It can be provided in appropriations acts, authorization acts, or in
permanent law.

5Throughout this report we use the term "mortgage payment" to mean an
amount equal to the interest and amortization on an acquired capital
asset.

would impute an annual capital usage charge on those assets to using
agencies. The CAF would collect these charges on existing capital and
forward them to Treasury; they could not be used for future projects. See
figure 1 for an example of the current capital financing process compared
to the process under a CAF mechanism.

    Figure 1: Up-Front Financing of Federal Capital Assets before and after
                Establishing CAFs: New Asset Obtained in Year 1

Source: GAO.

aThis example assumes the asset is acquired at the beginning of the fiscal
year and rented for the entire year. The mortgage payment equals
approximately $644,000, assuming a $10 million loan for 30 years at a 5
percent interest rate with monthly payments.

This new system for financing assets would increase management and
oversight responsibilities for Treasury, departments with CAFs, OMB, and
CBO. Treasury would have to set up CAF accounts and manage and report on
the borrowing authority and mortgage payments; it would consider charging
an administrative fee to cover the costs of these new responsibilities.
Departments would have to account for the transactions between the CAF,
Treasury, and the subunits, and perform oversight responsibilities. OMB
would have to issue guidelines to determine the types of assets to include
in the CAF and the method for calculating a capital usage charge for
existing capital. OMB and Congress would have oversight responsibilities,
which would need to be spelled out. In addition, OMB and CBO would have to
determine the scoring of both the initial up-front borrowing authority and
the subsequent appropriations for the annual mortgage payments. Although
on a gross basis the BA would be appropriated twice, the payments are
purely intragovernmental and offset one another so no adjustments would be
needed to any appropriation subcommittee allocations6 for new assets. If
existing capital assets are included in a CAF, the scoring of annual
capital usage charges would have to be determined.

CAFs might improve decision making by allocating capital costs to
programs. This could be part of a broader effort to include full costs in
program budgets as a way of facilitating comparison across programs. CAFs
might also ameliorate the spikes in BA for subunits that are sometimes
said to inhibit capital acquisition. However, we found that agencies are
implementing or using alternative mechanisms to address these challenges.
For example, some agencies are developing and beginning to use asset
management systems that should allow them to evaluate and record the
condition of existing assets, estimate the funding required to sustain
capital assets over time, and prioritize maintenance and improvements.7
Other agencies are using full cost information from accounting systems to
aid in budget decisions, although more progress is

6In this report, subcommittee allocations refer to the distribution by the
House and Senate appropriations committees of total new BA and outlays to
the 13 appropriations subcommittees as required by the Congressional
Budget and Impoundment Control Act of 1974. This allocation limits the
total budget authority and outlays available for all accounts under the
subcommittees' jurisdiction.

7We describe some of the asset management systems in this report and in
GAO, Budget Issues: Agency Implementation of Capital Planning Principles
Is Mixed, GAO-04-138 (Washington, D.C.: Jan. 16, 2004). However, it was
beyond the scope of both these reports to evaluate the effectiveness of
the systems described.

needed before their cost accounting systems can fully inform decision
making, including capital planning and budgeting.

Regarding spikes in BA, the case study subunits we spoke with generally
did not consider them to be an impediment to acquiring needed capital.
They explained that spikes are sometimes created by the changing
priorities and funding decisions of Congress rather than agency budget
requests. In other cases, agencies avoid spikes by funding capital in
useful segments or using no-year authorities.8 Both of these approaches
allow an agency to spread the total cost of an asset over time by either
completing a project in phases or accumulating numerous years of funding
in no-year accounts until the total amount of up-front BA has been
appropriated. In addition, agency working capital funds (WCF)9 and the
Federal Buildings Fund (FBF)10 at GSA are being used to finance some
capital assets. By using these funds, which charge a rent payment or user
fee, capital costs are allocated to programs and spikes in BA are averted.

In addition to describing alternative methods for achieving CAF goals,
many of those we spoke with voiced concerns about CAFs. Case study agency
and Treasury officials raised questions about the long-term feasibility of
making fixed annual mortgage payments, especially in times of constrained
budgets. Some subunits and congressional appropriations staff were
concerned about shifting more control over capital assets from the subunit
to the department level. In general, almost all agency officials,
congressional staff, and budget experts that we interviewed concluded that
CAFs sounded complicated and many questioned whether the challenges that
CAFs are designed to address were problematic enough to warrant

8No-year authority refers to budget authority that remains available for
obligation for an indefinite period of time, usually until the objectives
for which the authority was made available are attained.

9A working capital fund is a revolving fund that operates as an accounting
entity. The assets are capitalized and all income is in the form of
offsetting collections derived from the funds' operations and available in
their entirety to finance the funds' continuing cycle of operations
without fiscal year limitation.

10The FBF is a governmentwide revolving fund established in 1972 and is
the principal funding mechanism for the Public Buildings Service (PBS).
Within GSA, PBS leases office space to federal customer agencies. PBS
collects rent from federal tenants, which is deposited into the FBF.
Congress exercises control over the FBF through the appropriations process
that sets annual limits on how much of the fund can be expended for
various activities. In addition, Congress may appropriate additional
amounts for the FBF.

their adoption; especially given the additional budget complexities CAFs
create.

We identified several other pertinent issues to weigh when considering the
implementation of CAFs. Perhaps the most difficult issue to tackle would
be the implementation of an annual capital usage charge on existing
capital assets. The argument in favor of these charges is that they would
allow programs to show the full annual cost of capital in their budgets
and in doing so establish a level playing field for federal capital
investment to allow for comparisons across programs. The annual capital
usage charge also might influence agency managers to dispose of excess
capital assets. If existing capital is not included, it would be decades
before all programs showed the full annual cost of capital in their
budgets, and programs purchasing new capital would appear more expensive
than those using existing capital. However, there is no agreement on the
need for such a charge or how it would be computed.

Beyond that, although cost allocation efforts may increase the
transparency of total program costs, it is not clear that CAFs would
really create incentives for managers to make better decisions about new
or existing capital assets, especially if annual mortgage payments and
capital usage charges are automatically included in subunit
appropriations. For existing capital assets, mission responsibilities,
legal requirements for federal property sales, and financial costs would
likely constrain an agency's ability to dispose of surplus assets. If
implemented, CAFs would create a whole new set of circumstances that would
need to be addressed if an asset (which would be owned by the CAF) is sold
or transferred-how would a "sale price" be determined between federal
agencies, how would the full repayment of CAF debts be ensured, and how
would the agencies' budgets be adjusted? Finally, even if guidelines
indicate that most capital assets should be included in the CAF, it is
likely that some capital assets would continue to be funded through
currently existing mechanisms such as WCFs, FBF, and the Federal
Technology Service's (FTS) Information Technology (IT) Fund.11

11The IT fund is a full-cost recovery revolving fund that provides federal
agencies with IT products and services. FTS recovers both the costs of
products and services and the costs of their delivery through the IT Fund.
The IT Fund was authorized by the Paperwork Reduction Reauthorization Act
of 1986 as included in section 821(a)(1) of Public Law 99-500 and 99-691;
40 U.S.C. 322.

In the past, we saw some merit in the broad concept behind CAFs12 and
agreed that CAFs should be explored as a financing mechanism. After having
done an in-depth examination of the specifics and after discussing the
perspectives of others who would be involved in implementing CAFs, we now
see that implementation challenges are significant and that agencies have
adopted other mechanisms to address challenges CAFs were designed to
address. It is clear that in a mechanical sense CAFs could operate as
intended, albeit by increasing the complexity of capital asset financing.
However, we found there is little assurance that this increased complexity
will better achieve what is already being accomplished through the
alternative mechanisms discussed in this report. Without stronger
justification and a clear plan for handling the potential difficulties
raised in this report, CAFs would absorb the time and energy of those
involved in budgeting for capital without commensurate benefit.

We obtained comments on a draft of this report from OMB, Treasury, GSA and
our case study agencies-USDA and DOI. Treasury, GSA, USDA and DOI
generally agreed with the report. Treasury, USDA, DOI and OMB provided
technical comments, which have been incorporated as appropriate. OMB
provided oral comments and agreed with our description of the mechanics of
CAFs and concurred that spikes in BA for capital assets could be
alleviated through other means. OMB also recognized the problems with CAFs
that are highlighted in this report, including those related to existing
capital, and agreed that the complications of designing and operating CAFs
might outweigh the benefits. However, they disagreed with our description
of the goals of CAFs because they view CAFs as part of a broader effort to
have program budgets reflect full annual costs in order to change
incentives for decision makers. They do not believe alternative mechanisms
achieve that goal.

We recognize that if the sole or primary purpose of a CAF is to embed
costs in the program budgets, then the alternatives discussed in this
report do not achieve that purpose. However we believe, as highlighted in
the Report of the President's Commission to Study Capital Budgeting, that
the primary goal of CAFs is to improve decision making for capital. Asset
management and cost accounting systems, when fully implemented, provide
invaluable information that will assist decision makers in

12GAO, Accrual Budgeting: Experiences of Other Nations and Implications
for the United States, GAO/AIMD-00-57 (Washington, D.C.: Feb. 18, 2000).

determining how much and what types of capital are needed. While this
information may not necessarily be reflected in program budgets, it is
available to aid in budget and program decision making. The fact that many
of these systems are in relatively early stages of development also
increases our concern about CAFs. In a recent report, we noted the belief
among some agency officials, congressional appropriations committee staff,
and budget experts that improving underlying financial and performance
information should be a prerequisite to efforts to restructure program
budgets.13 We argue this would also be true for CAFs, since without
adequate measures of program costs and an ability to identify capital
priorities, a new financing mechanism would do nothing to address the
basic challenges of determining how much and what types of capital are
needed. Moreover, we are not convinced that CAFs and the annual mortgage
payments they would require would change incentives for program managers
or other decision makers, especially if annual mortgage payments and
capital usage charges are automatically included in subunit
appropriations. In conclusion, we remain of the view that the operational
challenges of CAFs outweigh the benefits and that alternative mechanisms
described in this report can more simply promote improved decision making.

Background	The federal government acquires a wide variety of capital
assets for its own use including land, structures, equipment, vehicles,
and information technology. Large sums of taxpayer funds are spent on
these assets, and their performance affects how well agencies achieve
their missions. To directly acquire an asset, agencies generally are
required to have full upfront BA for the total asset cost-usually a
sizable amount.14 This requirement allows Congress to recognize the full
budgetary impact of capital spending at the time a commitment is made;
however, it also means that the full cost of an asset must be absorbed in
the annual budget of an

13GAO, Performance Budgeting: Efforts to Restructure Budgets to Better
Align Resources with Performance, GAO-05-117SP (Washington, D.C.: February
2005), 15.

14While complying with up-front funding, agencies may choose to structure
capital purchases into a series of useful segments. A useful segment of a
capital project is a component that either (1) provides information that
allows the agency to plan the capital project, develop the design, and
assess the benefits, costs, and risks before proceeding to full
acquisition (or canceling the acquisition) or (2) results in a useful
asset for which the benefits exceed the costs even if no further funding
is appropriated.

agency or program, despite the fact that benefits may accrue over many
years. This up-front funding requirement has presented two challenges for
capital planning and budgeting at the federal level.

One challenge is how to permit "full cost" analysis and to promote more
effective capital planning and budgeting by allocating capital costs on an
annual basis to programs that use capital. Allocating capital costs over
the assets' useful lives ensures that the full annual cost of resources a
program uses is considered when evaluating the program's effectiveness. It
can make program managers more aware of on-going capital costs, thus
promoting more effective decision making for capital. It may also
contribute to equalizing comparisons across different programs or
different approaches to achieving similar goals.

A second challenge is how to address the possible bias against the
acquisition of necessary capital assets that may be created by spikes
(large, temporary, year-to-year increases in BA), which can make capital
assets seem prohibitively expensive in an era of resource constraints. GAO
has reported in the past that agencies view up-front funding as an
impediment to capital acquisition because of the resulting spike in BA.15
CAFs have been suggested as a capital asset financing approach that would
benefit federal departments and their subunits by addressing both of these
challenges. CAFs would be department-level funds that use annually
appropriated authority to borrow from the Treasury to purchase
federallyowned assets16 needed by subunits of the department. These
subunits would then pay the CAF a mortgage payment sufficient to cover the
principal and interest payment on the Treasury loan. The CAF would use
those receipts only to repay Treasury and not to finance new assets.

15GAO, Budget Issues: Budgeting for Capital, GAO/AIMD-97-5 (Washington,
D.C.:
Nov. 12, 1996); Budget Issues: Alternative Approaches to Finance Federal
Capital,
GAO-03-1011 (Washington, D.C.: Aug. 21, 2003); and Capital Financing:
Partnerships and
Energy Savings Performance Contracts Raise Budgeting and Monitoring
Concerns,
GAO-05-55 (Washington, D.C.: Dec. 16, 2004).

16It would not be appropriate or useful to include in the CAF grants to
states or localities for what, in other contexts, may be deemed to be
capital expenditures, such as those for highways. The grant itself is the
program; highways and other federally assisted capital assets are not
owned by the federal government and are not being used by the federal
government in its own operations, so there are no federal programs to
which the cost of using this capital should be allocated for budget
decision making.

The CAF concept was formally proposed in the February 1999 Report of the
President's Commission to Study Capital Budgeting17 as a mechanism that
would help improve the process by which annual budget decisions are made
by promoting better planning and budgeting of capital expenditures for
federally owned facilities. The report states that by ensuring that
individual programs are charged the true cost of using capital assets, the
CAF encourages managers to make more efficient use of those assets. The
Commission report also argues that CAFs could help smooth out the spikes
in BA experienced by subunits with capital project requests. By
aggregating all up-front BA for capital requests at the department level,
subunit budgets would reflect only an annual payment for capital. Since
the Commission report, CBO, GAO, and the National Research Council (NRC)
have all agreed that CAFs should be explored as a capital financing
mechanism.18

CAFs were also discussed in the President's Fiscal Year 2004 Budget issued
in February 2003. 19 The section on "Budget and Performance Integration"
briefly described the concept and reports that draft legislation creating
CAFs has been developed, discussed with agencies, and improved. It said
that CAFs would be one way to show the uniform annual cost for the use of
capital without changing the requirement for up-front appropriations. At
this time, OMB's interest in CAFs appears to have waned. CAFs were not
mentioned in the President's Budget in either fiscal year 2005 or 2006 and
the CAF legislation described has not been introduced.

Scope and	To address our objectives, we reviewed the available literature
describing the CAF concept. We also interviewed budget experts at OMB and
CBO to

Methodology	gain a more thorough understanding of how CAFs would operate
and discuss issues involved with their implementation. This permitted us
to describe a theoretical CAF with some operational detail. Additionally,
we

17The President's Commission to Study Capital Budgeting, Report of the
President's Commission to Study Capital Budgeting (Washington, D.C.:
February 1999).

18CBO, The Budgetary Treatment of Leases and Public/Private Ventures
(Washington, D.C.: February 2003); GAO/AIMD-00-57; NRC, Investments in
Federal Facilities: Asset Management Strategies for the 21st Century
(Washington, D.C.: The National Academies Press, 2004).

19OMB, Analytical Perspectives, Budget of the United States Government,
Fiscal Year 2004 (Washington, D.C.: 2003), 13.

sought the views of the many parties that would be affected if CAFs were
established. Since agency and congressional officials were generally
unaware of the CAF concept, we developed a brief summary describing the
general mechanics of a CAF and shared that summary prior to interviews in
order to generate discussion.

To get the department perspective, we chose USDA and DOI as case studies.
Both of these departments have substantial and varied capital needs.20
Capital assets acquired by USDA and DOI include land, buildings, research
equipment, laboratories, quarantine facilities, dams, bridges, parklands,
roads, trails, vehicles, aircraft, and information technology (hardware
and software). In addition, each department has multiple subunits that use
capital assets to achieve their missions-important for examining the
question of subunit spikes. We interviewed officials at the department and
subunit levels to gather their opinions and insights on the operation,
benefits, and difficulties of CAFs. Specifically within USDA we spoke with
officials in the Animal and Plant Health Inspection Service (APHIS), the
Agricultural Research Service (ARS), and the Forest Service (FS). Within
DOI, we spoke with officials in the National Park Service (NPS) and the
Bureau of Land Management (BLM). During these discussions, agency
officials also compared CAFs (as described in our summary) with current
practices used for planning, budgeting, and acquisition of capital assets.

Since congressional approval would be necessary for the creation and
operation of CAFs, we spoke with staff on the House and Senate Budget
Committees, the House and Senate Appropriations Subcommittees on the
Interior, and the House Appropriations Subcommittee on Agriculture21 to
get their opinions on the proposed CAF mechanism. We also interviewed
officials at Treasury, which would be responsible for managing the
borrowing authority. In addition, we spoke with officials at GSA to
discuss how a CAF might affect the FBF, used by some federal agencies to
acquire

20We used character class data from OMB's MAX system to identify
departments with substantial capital budget authority over the last 12
years. These character classes include Construction and Rehabilitation
(1312 and 1314), Major Equipment (1322 and 1324), and Purchases and Sales
of Land and Structures (1340). Major equipment includes capital purchases
of information technology but excludes the support services related to
information technology purchases. MAX is the computer system used to
collect and process information needed to prepare the President's Budget.

21We were unable to meet with staff from the Senate Appropriations
Subcommittee on Agriculture.

federal office space and the FTS, used by some federal agencies to acquire
IT.

Finally, we reviewed agency documents including asset management plans,
accounting system descriptions, capitalization policies, and working
capital fund information. We also examined our prior work, financial
accounting standards, and various legal and budgetary sources specifically
related to federal property management.

We recognize that our findings on agency perspective, which are based on
interviews with five subunits within two departments, may not be
applicable to all agencies within the federal government. However, we were
struck by the consistency in department and subunit reaction to the
concept, especially when followed by comparable reactions from
congressional officials. Our work was conducted in Washington, D.C., from
May 2004 through January 2005 in accordance with generally accepted
government auditing standards.

  CAF Operations Would Create a New Financing System and New Oversight
  Responsibilities

Implementing CAFs would change the current process for financing new
federal capital projects. In addition, if all existing capital assets of a
department and its subunits were transferred to the CAF, the CAF would
impute an annual capital usage charge on those assets to using agencies.
This additional complication could be avoided if CAFs were limited to new
assets. However, this would mean it would be decades before all programs
showed the full annual cost of capital in their budgets.

Although in many respects CAFs are accounting devices to record financial
transactions, their creation would create new management and oversight
responsibilities for many federal entities. Treasury would have primary
responsibility for administering the borrowing authority. Both Treasury
and those departments with CAFs would be required to keep track of the
many CAF transactions. The management and oversight responsibilities of
the departments would need to be clearly spelled out in order for CAFs to
operate effectively. OMB would likely have to issue guidelines on
operation specifics and OMB and the congressional appropriations committee
staff would have to review the CAFs to ensure they were operating
properly. OMB and CBO would score (estimate) the CAFs' and subunits'
BA-both the initial authority to borrow and the subsequent appropriations
used for repayment. The scoring of the annual capital usage charges, if
CAFs were applied to existing capital, has not yet been developed.

    CAFs Would Be Positioned at the Department Level and Create a More Complex
    Process for Financing Capital

Although CAFs do not currently exist, we can describe how they would
likely operate based on written proposals and our discussions with budget
experts. CAFs would be established at the department level as separate
accounts that would receive up-front authority to borrow (provided in
appropriation acts) on a project-by-project basis, for the construction
and acquisition of large capital projects for all of the subunits within a
department. For those departments with subunits split between two
appropriation subcommittees, it is likely that two CAFs would be
necessary. For example, DOI receives appropriations through two
subcommittees: the Energy and Water Development Subcommittee, which is
responsible for Bureau of Reclamation (Reclamation) programs; and the
Interior and Related Agencies Subcommittee, which is responsible for all
other Interior programs. CBO, OMB, and agency officials we spoke with
generally believed that having a CAF that crossed subcommittee
jurisdictions would create many problems, thus it would likely be
necessary for departments to have a separate CAF for each subcommittee
with which they work. Using the example above, DOI would have one CAF for
Reclamation and a second for the remaining subunits within DOI.
Alternatively, CAFs could be situated at the appropriation subcommittee
level rather than the department level, with each of the 13 subcommittees
appropriating to their respective CAF for the agencies under their
jurisdiction.22 Some congressional officials did not seem to think that
this would be the most effective arrangement and raised the point that
increased resources might be needed at the subcommittee level to manage
CAF transactions. In addition, OMB argued that CAFs should be located at
the department level because the department is the focus of accountability
for planning and managing programs and capital assets, as well as for
budget execution and financial reporting.

The CAF would receive appropriations for the full cost of an asset (or
useful segment of an asset) in the form of borrowing authority. Like all
BA, the borrowing authority for each CAF-financed project would specify
the purpose, amount, and duration of the authority. Unless the asset is to
be available for use in the same fiscal year, the subunit itself would
receive no appropriations. The CAF would use its authority to borrow from
the Treasury's general fund to acquire the asset for the subunit. When the
asset became usable, the subunit would begin to pay the CAF an amount
equal to

22The President's Commission to Study Capital Budgeting, Report of the
President's Commission, 33 and NRC, Investments in Federal Facilities, 82.

a mortgage payment consisting of interest and principle. These equal
annual payments would consist of the principal amortized over the useful
life of the asset and include interest charges at a rate determined by
Treasury (based on the average interest rate on marketable Treasury
securities of comparable maturity). The CAF would use these mortgage
payments to repay Treasury for the funds borrowed plus interest. Unlike a
revolving fund, the mortgage payments collected by the CAF would be used
only to repay Treasury and could not be used to finance new assets.

For each project funded through the CAF, the subunit's annual budget
request would need to include the annual mortgage payment in each year,
for the useful life of the asset (or until the asset was sold or
transferred). The subunit would need annual appropriations for these
payments, along with its other operating expenses. On the basis of our
discussions, we conclude that the appropriations from which the payments
are made would be discretionary as opposed to mandatory. They would not be
provided as a line item for mortgage payments to the CAF, but would be
part of the subunit's total appropriation. While the subunit would be
required to make the annual payment, there would be no guarantee that
Congress would include the additional amounts to cover the payment in the
subunit's appropriation.

At some point, the mortgage on an asset would be "paid off." However, if
annual capital usage charges on existing capital were established,
payments would continue, although the amount of the payments would depend
on the method used to calculate the charges for existing capital. Any
imputed charges collected by the CAF would be transferred to the general
fund of Treasury and not be available to finance new assets. Later in this
report we discuss in more detail the idea of imputing a capital usage
charge on existing capital.

    Treasury Would Oversee Borrowing Authority Used to Acquire Capital Assets

Treasury is responsible for administering and managing borrowing
authority. Treasury officials explained that within the department, the
Financial Management Service (FMS) would have responsibility for setting
up the accounts to correspond with each CAF created. Before a CAF could
actually borrow from Treasury, an agreement would have to be signed
establishing the interest rate and repayment schedule. Treasury officials
recommended that OMB establish guidelines to specify the useful life of
capital assets so departments would abide by an appropriate amortization
schedule and not attempt to lower payments by lengthening the asset's

useful life. The standards issued by the Federal Accounting Standards
Advisory Board (FASAB) on how to account for property, equipment, and
internal-use software could be useful in developing these guidelines.23
According to Treasury officials, FMS would also be responsible for
preparing the warrants, an official document that establishes the amount
of monies authorized to be withdrawn from the central accounts maintained
by Treasury, and would report annually on account activity. The Bureau of
Public Debt would have the most day-to-day interaction with the CAF. It
would be responsible for transferring the borrowed funds to the department
and for receiving payments. Although Treasury officials did not think it
would be an unmanageable task, they said that tracking individual
transactions could become complicated, depending on the level of detailed
reporting required, and would certainly require additional staff time. To
cover these costs, they would want to charge an administrative fee, as
they do for trust funds.

    CAFs Would Add Complications to Oversight and Scoring

A CAF is an additional layer of administration that could complicate
program management rather than streamline it. At the department level, the
chief financial officer would likely be responsible for the financial
operation of the CAF. Department heads would need to specify duties for
those with capital asset management and oversight responsibilities
according to the unique needs of the department. Oversight functions would
include accounting for all the transactions between the CAF and Treasury
as well as between the CAF and the subunits. In addition, the managerial
relationship between the CAF and individual subunits would have to be
worked out. OMB would also likely have new responsibilities. For example,
OMB would probably have to develop guidelines on issues such as (1) the
types of assets to include in the CAF, (2) the amortization schedule for
various types of assets, (3) the method for calculating a capital usage
charge on existing capital (along with CBO and Congress), and (4) the
relationship between a CAF and FBF.24 Indeed, the NRC report argued that
oversight and management of CAFs should actually reside at OMB. Although
OMB officials provided no details, they agreed that they would have some
responsibility for reviewing CAFs, as would congressional committees.

23Statement of Federal Financial Accounting Standards, No. 6, Accounting
for Property, Plant and Equipment and No. 10, Accounting for Internal Use
Software.

24All of these issues are discussed in more detail in other sections of
this report.

As they do for all appropriation actions, CBO and OMB would score the CAF
and subunit BA-both the initial authority to borrow and the subsequent
appropriations used for repayment. Although the net amounts of BA and
outlays for capital acquisitions would not change, the type of BA would.
Currently, annual appropriations, which allow program managers to incur
obligations and make outlays with no additional steps, are provided for
most capital acquisitions. A CAF, however, would be appropriated up-front
borrowing authority. On a gross basis, the BA would have to be
appropriated twice, once as up-front borrowing authority and incrementally
over time through appropriations for the annual mortgage payment. Since
the annual mortgage payment is purely intragovernmental, the subunit's BA
and outlays are offset by receipts in the CAF, so the total BA and outlays
are not double-counted. Therefore, appropriation subcommittee allocations
would not need to be adjusted if a CAF were used for new assets.

The initial borrowing authority would be equal to the asset cost and would
be scored up front in the CAF budget. When the annual mortgage payments
begin, the amount provided in the subunit's budget would equal the
mortgage payment and would be scored as discretionary BA. The mortgage
payment would then be transferred to the CAF and, as a receipt, be
considered mandatory BA. However, according to OMB, it would be treated as
a discretionary offset for scoring purposes.25 The payment and receipt
would completely offset each other within the appropriation subcommittees'
totals and in the BA and outlay totals for the federal budget as a whole.

When the CAF repays Treasury using the mortgage receipts, scoring would
follow the current guidelines for debt repayment transactions. The
mortgage receipt would be considered mandatory BA and be used to repay
Treasury; however, the portion of the mortgage payment that corresponds to
the amortization of the asset cost would be deducted from the BA (and
outlay) totals. When collections are used for debt repayment, they are
unavailable for new obligations, and therefore are not BA. If they were
counted, the BA and outlay totals would be overstated over the life of the
loan. According to OMB, the remaining mandatory BA would be obligated and
outlayed for interest payments to an intragovernmental receipt account in
Treasury, but would not be scored. At this time, the scoring of

25According to OMB officials, this treatment of budget authority can be
used for certain transactions.

annual capital usage charges on existing capital assets has not been
determined.

  CAF Benefits Can Be Achieved through Alternative Means Without the Added
  Budget Complexity

CAFs have been proposed as a way to address two challenges that arise from
the full up-front funding requirement for capital projects. The first
challenge is to facilitate program performance evaluation and promote more
effective capital planning and budgeting by allocating capital costs on an
annual basis to those programs using the capital. By having annual cost
information, managers can better plan and budget for future asset
maintenance and replacement. During our interviews, we learned that asset
management and cost accounting systems are currently being implemented
that could be used to address this problem. These systems are designed to
provide the information necessary for improved priority setting and better
decision making, although we found that many agencies are still working to
fully implement and use these systems. The second challenge-managing
periodic spikes in BA caused by capital asset needs-if considered a
problem at all, is managed by our case study agencies through existing
entities and practices, such as the use of WCFs. Consequently, CAFs appear
to offer few benefits over and above those provided by other mechanisms
being put into place or in use. In addition, officials at the department
and subunit level and key congressional staff we spoke with have a number
of concerns about adopting CAFs as an alternative financing method. Most
of those we spoke with said CAFs sounded like a complicated mechanism to
achieve benefits that can be achieved in simpler ways and some worried
that implementation of CAFs could distract from current efforts to improve
capital decision making.

    Allocating Annual Capital Costs and Improving Decision Making for Capital
    Assets May Be Achieved through Existing Initiatives

Officials we interviewed reacted to our presentation of the CAF mechanism
by describing current agency initiatives and existing mechanisms that they
believe can better achieve the ultimate goal of improving budgeting and
decision making for capital. We found that some agencies currently make
use of asset management plans to collect, track, and analyze cost
information and to assist management in budget decisions and priority
setting. Accounting systems that report full costs are also being
developed that will include the cost of capital assets in total program
costs and will provide a tool for agency managers to make better decisions
and use capital more efficiently. Once fully implemented, these methods
will provide agencies with the ability to assign costs at the program
level and link those costs to a desired result. The information provided
should lead

The Departments of the Interior and Agriculture Are Implementing Asset
Management Systems to Make Informed Decisions on Capital Investment

agencies to consider whether they will continue to need the current
quantities and types of fixed assets they own to meet future program
needs.

As we have reported in previous work, leading organizations gather and
track information that helps them identify the gap between what they have
and what they need to fulfill their goals and objectives.26 Routinely
assessing the condition of assets and facilities allows managers and their
decision makers to evaluate the capabilities of current assets, plan for
future asset replacements, and calculate the cost of deferred maintenance.
We found that asset management systems are being developed and implemented
at some agencies as a mechanism to aid in the identification of asset
holdings and prioritization of maintenance and improvements.

For example, we reported in 2004 that NPS, within DOI, is currently
implementing an asset management process.27 If it operates as planned, the
agency will, for the first time, have a reliable inventory of its assets,
a process for reporting on the condition of those assets, and a systemwide
methodology for estimating deferred maintenance costs. The system requires
each park to enter all of its assets and information on its condition into
a centralized database for the entire park system and to conduct annual
condition assessments and regular comprehensive assessments. This new
process will not be fully implemented until fiscal year 2006 or 2007, and
will require years of sustained commitment by NPS and other stakeholders.

26GAO, Executive Guide: Leading Practices in Capital Decision-Making,
GAO/AIMD-99-32 (Washington, D.C.: December 1998).

27GAO-04-138. In this report we also found that three of four case study
agencies-NPS, the Department of Veterans Affairs, and the National Oceanic
and Atmospheric Administration-lacked current asset condition data.

Figure 2: NPS Asset Management Plan

Department officials provided us a prototype of an asset management plan
for the Grand Canyon National Park. The objective is to establish the
total cost of ownership of the Grand Canyon's asset inventory and to
provide a tool to aid managers in budget decisions, priority setting, and
communication. It focuses on four key questions about the following:

o  What inventory NPS owns in the park.

o  The condition of assets.

o  Current replacement values.

o  What operations and resources are required to properly sustain the
asset inventory. Information on the park's inventory is gathered from
condition assessments, operations and maintenance budgets, staff
experience, and industry standard sources.a This information helps clarify
asset maintenance and operations requirements, which are then compared to
agency budget data to determine if funding levels are adequate to sustain
the capital over time. In addition, managers use a facility condition
index plotted against an asset priority indexb to restore assets in
priority order and identify assets for disposal.

Source: DOI.

aIndustry standard sources refer to the facility condition index, which is
considered to be a leading metric for assessing asset conditions. It is
calculated by dividing the total project requirements by the replacement
value of the asset.

bThe asset priority index is a score that park leadership assigns and is
reflective of the asset's relevance to carrying out the park mission.

According to NPS documents, this approach and the information captured in
the asset management plan provides Grand Canyon National Park managers
with the knowledge and specifics to make informed capital investment
decisions and to develop sound business cases for funding requests. The
appropriators for NPS that we spoke with agreed that the additional
funding they have provided for condition assessments and asset management
has improved planning and decision making at NPS. Department officials
told us that these types of asset management plans would eventually be
completed for all capital-holding subunits within DOI. The completion of
this management system is especially important for DOI because much of its
mission is the upkeep and improvement of its capital for use by the
public.

FS, whose capital includes numerous trails, roads, and recreation
facilities, has implemented and is continuing to enhance its asset
management system referred to as Infrastructure (INFRA). INFRA has been in
production since 1998 and served as the agency's primary inventory
reporting and portfolio management tool for all owned real property until
May 2004. FS officials said that they have used INFRA to assist management
in prioritizing backlogs of maintenance and renovations. According to
these officials, INFRA allows for the transfer of FS asset

inventory data directly into USDA's asset inventory system known as the
Corporate Property Automated Information System (CPAIS). CPAIS, which
agency officials said was modeled after INFRA and further enhanced to
include leased property and GSA assignments, was implemented in May 2004
and maintains data elements necessary to track and manage owned property,
leased property, GSA assignments, and interagency agreements. The system
will provide the department and its subunits with the capability to
increase asset utilization and cost management and to analyze and reduce
maintenance expenses. The primary users of the system are those subunits
with considerable capital needs, according to agency officials. ARS's
capital is mostly high-priced laboratories, specific scientific equipment,
and research facilities, and officials are confident that CPAIS will
provide the information needed to ensure accountability over its real
property. ARS also has its own facilities division made up of contractors
and engineers that are equipped with the experience and expertise to
manage and oversee their specialized capital projects. APHIS officials
said they are in the process of doing facility condition assessments and
hope to use the information in order to better align its mission with its
strategic plan.

The need for asset management systems to aid agency officials in making
informed decisions was underscored in our report designating federal real
property as a new high-risk area in 2003. The report highlighted the fact
that in general, key decision makers lack reliable and useful data on real
property assets.28 In February 2004, the President issued an Executive
Order for Federal Real Property Asset Management. The order requires
designated agencies to have a real property officer and to implement an
asset management planning process. Its purpose is to promote the efficient
and economical use of America's real property assets and to assure
management accountability for implementing federal real property
management reforms.

28GAO, High-Risk Series: Federal Real Property, GAO-03-122 (Washington,
D.C.: January 2003).

Some Agencies Are Beginning to Use Full Cost Information to Make Budget
Decisions, Although Much Work Needs to Be Done

We found that some agencies are currently implementing cost accounting
methods, such as activity-based costing (ABC), to help determine the full
cost of a product or service, including the annual cost of capital, and
using that information to make budgeting decisions.29 For example, BLM has
implemented a management framework that integrates ABC and performance
information. We have previously reported that BLM's model fully
distributes costs and can readily identify, among other things, (1) the
full costs of each of its activities and (2) what it costs to pursue each
of its strategic goals. The system provides detailed information that
facilitates external reporting and can be used for internal purposes, such
as developing budgets and analyzing the unit costs of activities and
outputs.30 Integrating cost and performance information into one system
helped BLM become a finalist for the President's Quality Award in 2002 in
the "performance and budget integration" category. The bureau was
recognized for implementing a disciplined approach that allows it to align
resources, outputs, and organizational goals, and can lead to insights to
reengineer work processes as necessary. Among the results of its ABC
efforts, BLM has reported increased efficiency and success in completing
deferred maintenance and infrastructure improvement projects. BLM was at
the forefront of this cost management effort, which began in 1997 and has
now been adopted departmentwide as part of DOI's vision of effective
program management.

In another report, we described how the National Aeronautics and Space
Administration (NASA) is beginning to use accounting information to help
make decisions about capital assets.31 NASA's "Full Cost" Initiative
involves changes to accounting, budgeting, and management to enhance
cost-effective mission performance by providing complete cost information
for more fully informed decision making and management. The accounting
changes allow NASA to show the full cost of related projects and
supporting activities while the "full cost" budgeting uses budget
restructuring to better align resources with its strategic plan. The
accounting and budgeting portions of the initiative support the

29Statement of Federal Financial Accounting Standards, No. 4, Managerial
Cost Accounting Concepts and Standards for the Federal Government,
recommends that federal entities report the full costs of outputs in
general-purpose financial reports.

30GAO, Bureau of Reclamation: Opportunities Exist to Improve Managerial
Cost Information and Cost Recovery, GAO-02-973 (Washington, D.C.: Sept.
20, 2002).

31GAO-05-117SP.

management decision-making process by providing not only better
information, but also incentives to make decisions on the most efficient
use of resources. For example, NASA officials credited "full cost"
budgeting with helping to identify underutilized facilities, such as
service pools-the infrastructure capabilities that support multiple
programs and projects. NASA's service pools include wind tunnels,
information technology, and fabrication services. If programs do not cover
a service pool's costs, NASA officials said that it raises questions about
whether that capability is needed. NASA officials also explained that when
program managers are responsible for paying service pool costs associated
with their program, program managers have an incentive to consider their
use and whether lower cost alternatives exist. As a result, NASA officials
said "full cost" budgeting provides officials and program managers a
greater incentive to improve the management of these institutional assets.
Although accounting changes alone are not sufficient to improve decision
making and management, it is clear from discussions with NASA officials
and agency documentation that the move to full costing is a critical piece
of the initiative.

Some agencies still need to make more progress before their cost
accounting can more fully inform their decision making, including
decisions on capital planning and budgeting. In a 2003 report looking at
the financial management systems of 19 federal departments, we found that
although departments are required to produce information on the full cost
of programs and projects, some of the information is not detailed enough
to allow them to evaluate programs and activities on their full costs and
merits.32 For example, the Department of Defense (DOD) does not have the
systems and processes in place to capture the required cost information
from the hundreds of millions of transactions it processes each year.
Lacking complete and accurate overall life-cycle cost information for
weapons systems impairs DOD's and congressional decision makers' ability
to make fully informed decisions about which weapons, or how many, to buy.
DOD has acknowledged that the lack of a cost accounting system is its
largest impediment to controlling and managing weapon systems costs. Our
report states that departments are experimenting with methods of
accumulating and assigning costs to obtain the managerial cost information
needed to enhance programs, improve processes, establish fees, develop
budgets, prepare financial reports, make competitive sourcing decisions,

32GAO, Financial Management: Sustained Efforts Needed to Achieve FFMIA
Accountability, GAO-03-1062 (Washington, D.C.: Sept. 30, 2003).

and report on performance. As departments implement and upgrade their
financial management systems, opportunities exist for developing cost
management information as an integral part of the systems to provide
important information that is timely, reliable, and useful.

CAFs Might Smooth Budget The President's Commission to Study Capital
Budgeting and NRC have Spikes, but Benefit May Be suggested that a CAF
might help ameliorate the spikes in agency budgets Minor that often result
from large periodic capital requests by smoothing capital

costs over time and across subunits. Our analysis of recent trends in BA
for capital acquisitions clearly shows the presence of spikes at the
subunit level. See figure 3 for an illustration of budget spikes and
potential smoothing effects of a CAF at ARS.

Figure 3: Illustration of Budget Spikes and Potential Smoothing Effects of
a CAF at ARS

Percent change in budget authority for capital assets 150

125

100

75

50

25

0

-25

-50

-75

-100 1 2 3 4 5 6 7 8 9101112131415161718192021222324252627282930

Year

                    Up-front financing under current system

                      Annual mortgage payment under a CAF

Source: GAO analysis of OMB MAX data.

Note: In this figure, the first 12 years of data are based on actual BA
for capital assets for fiscal years 1994 through 2005. To simulate a
long-term trend, we replicate this data for years 13 through 30. A 20-year
repayment term is used to calculate the annual mortgage payment, which
does not include an interest charge. We obtained BA data for capital
assets from OMB's MAX database as described in the Scope and Methodology
section.

However, these spikes did not appear to be a major concern to the case
study subunits we spoke with nor did they consider them a barrier in
meeting capital needs. Given current practices for financing capital
assets, it seems that some program managers and Congress have found ways
to cope with spikes in the absence of CAFs. As a result, the benefit of
smoothing costs with a CAF would be minimal.

Some Spikes May Be Created by Congressional Funding Decisions

Our prior work indicates that some agencies have complained that large
spikes in their budget hinder their ability to acquire the needed funding
to complete capital projects33 and reveals that some agencies have turned
to alternative financing mechanisms, such as incremental funding,
operating leases, and public-private partnerships, that allow them to
obtain assets without full, up-front BA.34 A few agency officials we spoke
with said that because of the up-front funding requirement, they have
sometimes opted for operating leases instead of capital leases or
constructing buildings. Operating leases are generally more expensive than
construction, purchase, or capital leases for long-term needs but do not
have to be funded up front. Nevertheless, the agencies we spoke with
reported that spikes are often created by the changing priorities of
Congress and its willingness to provide up-front funding for favored
capital projects. For example, ARS officials reported that appropriators
have increased the agency's budget in a given year to fund a new or
expanded facility that the subcommittee considered a priority.
Historically, the appropriations subcommittee for ARS (and all USDA
agencies except FS) has been active in initiating capital projects and
following through with the up-front funding necessary to build or acquire
assets. From ARS's perspective, budget spikes are not problematic because
of the perceived ease in obtaining needed funds. DOI also reported that
some of its subunits have received "waves" of funding for capital projects
largely dependent upon the priorities of Congress and the President.
Within DOI, BLM officials agreed that budget spikes were mostly a result
of congressional add-ons. On the other hand, NPS reported that most of its
capital projects are just not large enough to cause a noticeable budget
spike.

Staff from the congressional budget committee suggested that deliberations
during the appropriations process result in some smoothing at the
subcommittee level. The smoothing effects may not be apparent to agencies
when they review their individual budgets, but they are evident from a
governmentwide perspective. Historical analysis shows that federal
nondefense capital spending has remained relatively constant over the past
30 years.35

33GAO/AIMD-97-5.

34GAO, Budget Issues: Incremental Funding of Capital Assets, GAO-01-432R
(Washington, D.C.: Feb. 26, 2001) and GAO-03-1011.

35OMB, Historical Tables, Budget of the U.S. Government, Fiscal Year 2005
(Washington, D.C.: 2004), Table 9.3, 160-161.

Spikes Are Being Managed by Funding Useful Segments or Using No-Year
Authority

WCFs and FBF Can Be Used Both to Finance Capital Assets Without Spikes and
to Allocate Capital Costs

When spikes might be a problem, the departments and subunits we spoke with
have been able to manage them by dividing projects into useful segments
and accumulating funds with no-year authority. USDA and FS reported that
they have broken capital projects into useful segments and requested the
funding accordingly to minimize dramatic fluctuations in capital costs.
For example, USDA is currently renovating its headquarters in Washington,
D.C., and is using funds the department receives every other year to
finance the overhaul of one discrete section of the building at a time.
APHIS and BLM have also broken up large projects by funding the survey and
design phases in the first year and requesting funds for construction in
subsequent years. In addition, ARS and APHIS have authorities that allow
them to accumulate a specified amount or percentage of unobligated funds
until the amount is sufficient to cover the full up-front costs of the
desired asset. For example, ARS is building an animal health center in
Iowa, which costs an estimated $460 million. ARS received $124 million in
fiscal year 2004 towards the project and can accumulate that money in its
no-year account until the total amount to cover the costs is collected. In
its efforts to consolidate field offices, APHIS officials told us they
were granted authority to convert $2 million in unobligated balances into
no-year money each year for 3 years. The $6 million it was able to
accumulate allowed it to fund the consolidation with up-front funding. The
bureau hopes to expand this authority to apply to other capital, including
helicopters and airplanes.

WCFs, a type of revolving fund,36 are a mechanism that can be used both to
spread the cost of capital acquisition over time and to incorporate
capital costs into operating budgets. As reported previously,37 we found
that WCFs can be effective for agencies with relatively small, ongoing
capital needs because the WCFs, through user charges, spread the cost of
capital over time in order to build reserves for acquiring new or
replacement assets. Also, WCFs help to ensure that capital costs are
allocated to programs that use capital by promoting full cost accounting.
Since WCFs are designed to be self-financing, the user charges must be
sufficient to recoup the full cost of operations and include charges, such
as depreciation, to help fund capital replacement.

36Revolving funds are accounts authorized to be credited with collections
that are earmarked to finance a continuing cycle of business-type
operations without fiscal year limitation.

37GAO-97-5.

Some we spoke with use WCFs to finance capital assets such as IT
initiatives and equipment. For example, USDA's WCF provided funds to the
National Finance Center, one of its activity centers, to purchase and
implement a financial system. Department officials explained that after
the system became operational, the Finance Center charged the 28 user
entities a depreciation expense to recoup the costs of purchasing the
system so it could repay the WCF. In another example, the FS's WCF
purchases radio equipment, aircraft, IT, and other motor-driven equipment.
The equipment is rented out to administrative entities within the agency,
such as the National Forests and Research Experiment Stations, and to
outside agencies for a charge that recoups the costs of operation,
maintenance, and depreciation. The user charge is adjusted to include
sufficient funds to replace the equipment. Agency officials would like to
expand the WCF beyond just equipment and establish a facilities
maintenance fund. Through this fund, they would apply a standard charge
per square foot plus a replacement cost component. The charges would be
used for ongoing maintenance and replacement and they believe would help
influence line officers to reexamine capital needs. BLM's WCF functions
similar to that of the FS's WCF. BLM's WCF purchases vehicles, then
charges fees to users of the vehicles and uses the revenue to buy
replacement vehicles. In both of these examples, the WCF is designed to
accumulate the funds to absorb the up-front costs of the capital while the
user entities incur the annual costs of using the capital.

This mechanism operates similarly to a CAF, but with more flexibility in
the funding requirements. First, since WCFs are revolving funds, they
allow agencies to purchase new capital without a specific congressional
appropriation whereas a CAF would require a new appropriation to purchase
new capital. Second, WCFs are not subject to fiscal year limitations (they
have no-year authority) while CAFs would have project-by-project borrowing
authority specified in appropriation acts. Third, WCFs reflect annual
capital costs through a depreciation charge whereas CAFs would reflect
this cost through an annual mortgage payment of principal and interest.38
Hence, both would reflect the annual cost of capital in the subunits'
budgets.

38OMB noted that these two costs are not identical since CAF mortgage
payments would reflect an interest cost whereas WCFs do not reflect an
interest cost. However, besides depreciation, some WCFs are able to charge
additional amounts for future asset replacement.

To obtain federal office space, many agencies lease from and make rental
payments to GSA, which deposits those funds into the FBF. Although leasing
is recognized as being more expensive in the long run than ownership, some
agencies lease because it does not require as much up-front funding as
ownership (i.e., to avoid spikes). Although a CAF is conceptualized to
reduce the amount of up-front funding needed by subunits when acquiring
capital assets (while still requiring up-front funding at the department
level), it is not clear that having a CAF would encourage subunits to
build rather than lease office space. Two agency officials we spoke with
said that they would likely continue leasing and one commented that if
planning outright ownership, it would be easier to deal with obtaining the
traditional up-front funding than worry about the annual mortgage payments
required by a CAF. Through their charges, both WCFs and FBF spread the
cost of capital over time and ensure that capital costs are properly
allocated to the user programs.

    Agency Officials, Congressional Staff and Other Key Players Have Numerous
    Concerns About CAFs

Concerns over Receipt of Annual Mortgage Payment

Agency officials, congressional staff, and other key players raised
numerous concerns about CAFs. For example, department and subunit
officials are concerned that there is no guarantee or assurance that the
annual mortgage payments to be collected by the CAF will be adequately
funded in annual subunit appropriations. In addition, some subunits and
appropriators are reluctant to shift more control for capital planning and
budgeting to the department level. Congressional staff also raised
concerns about the feasibility of the congressional mind shift that would
be required to fund capital through a mechanism such as a CAF, especially
if a charge on existing capital is included, and questioned the value that
a CAF would really add to agency planning and budget decision making that
could not be obtained through other means. CBO and GSA were also
apprehensive and cautious about the usefulness of the CAF concept when
operating details were described in full. Most budget experts and agency
officials we spoke with agreed that the complexities involved in operating
a CAF would likely outweigh the possible benefits. A few worried that CAFs
might even divert attention from the current initiatives under way to
improve asset management and full costing.

Treasury, which would assume responsibility for collecting debt
repayments, was concerned that there would be no guarantee that future
appropriations would finance the mortgage payments, nor would there be any
enforcement mechanism by which Treasury could enforce repayment. Treasury
officials feared that over time other types of spending would take

priority over debt repayment. They based their concerns on the record of
some other programs that have struggled to repay debt or for which debt
has been "forgiven" or otherwise excused. For example, the Black Lung
Disability Trust Fund, which provides disability benefits and medical
services to eligible workers in the coal mining industry, has growing debt
and will never become solvent under current conditions.39 Although Black
Lung Disability Fund revenues are now sufficient to cover current
benefits, they do not cover either repayment of the over $8 billion owed
the Treasury or interest on that debt. Another example is the Bonneville
Power Administration (BPA), which is a federal electric power marketing
agency in the Pacific Northwest with authority to borrow from Treasury on
a permanent, indefinite basis in amounts not exceeding $4.45 billion at
any time. BPA finances its operations with power revenues and the loans
from Treasury, and has authority to reduce its debt using "fish credits."
This crediting mechanism, authorized by Congress in 1980, allows BPA to
reduce its payments to Treasury by an amount equal to mitigation
measures40 funded on behalf of nonpower purposes, such as fish mitigation
efforts in the Columbia and Snake River systems. BPA took this credit for
the first time in 1995 and has taken it every year since that time. The
annual credit allowed varies, but has ranged between about $25 million and
$583 million, including the use in 2001 and 2003 of about $325 million
total unused "fish credits" that had accumulated since 1980.

Some officials at the department and subunit level also raised concerns
about the long-run feasibility of fulfilling their mortgage payments over
the entire repayment period given that the payments are made from their
annual appropriations, which they expect to become increasingly
constrained. The mortgage payments would be relatively uncontrollable
items within an agency's budget, to the detriment of other, more
controllable items, such as personnel costs. Because the mortgage costs
would not change unless the asset is sold, managers would have less
flexibility in making budgeting decisions within stagnant or possibly
declining annual budgets that occur in times of fiscal restraint. BLM
officials said this type of fixed obligation, which could consume an
increasing share of its budget, could hinder its ability to address
emergency needs that arise during the year. For example, they cited a case
in which the

39The debt resulted from advances originally obtained to cover benefit
payments that coal taxes, the primary source of fund revenues, could not
provide.

40BPA is required by law to mitigate the impacts to fish and wildlife to
the extent they are affected by the construction and operation of the
Federal Columbia River Power System.

agency reprogrammed resources to deal with a landslide that occurred on
the Oregon coast in late 2003. BLM delayed other projects in order to
redirect funds for the removal and stabilization of the landslide and to
reopen Galice Creek Road, which is a major artery for public access,
recreation, and commercial activity such as timbering, as well as BLM and
FS administration. BLM officials questioned whether the fixed payment to
the CAF would constrain their ability to make adjustments such as this.
Many agency officials were skeptical of the idea that they could fulfill
annual mortgage payments to a CAF without squeezing program operations and
some said they would rather deal with the up-front funding requirement
than have to worry about annual mortgage payments.

The alternative to force-fitting a mortgage payment within agencies'
annual appropriations is to adjust agency budgets with an automatic add-on
equal to agencies' mortgage payments. While this would relieve budget
pressures at the agency level, it would probably not provide incentives or
influence managers to improve capital asset management and decision
making.

Concerns About Shifting More Under the CAF concept, requests for capital
projects would come from the

Control over Capital Assets to department level and the CAF would own all
capital assets. This would

the Department Level	shift more control of capital planning and decision
making from the subunit to the department level. Some agencies and one
appropriation subcommittee staffer said they would not favor this shift.
Several agencies feel that they have the expertise and experience to
better assess their own capital needs, which are often mission specific.
For example, ARS's capital consists of mostly scientific equipment,
laboratories, and research facilities designed for conducting agricultural
research in various climates. In fact, the agency has its own facilities
division consisting of contractors and engineers who are involved in the
management and oversight of capital projects. Similarly, APHIS's
facilities are mission specific. BLM's use of activity-based costing
allows it to assign capital costs to the program level and track those
costs to desired outputs. Consequently, the bureau has a more intimate
understanding of its capital needs and how capital contributes to carrying
out its mission. One agency raised the point that departmental management
might force bureaus to share facilities or later decide to use an asset
for purposes other than those originally intended. While some of these
departmental decisions might be beneficial, some agencies were skeptical
of departmental decision making.

Concerns over Problems Not Addressed, Additional Complexity, and Limited
Benefits

The officials we interviewed stated that there are important problems in
capital budgeting that CAFs do not address. Before the smoothing effects
of a CAF can be realized in the out years, the department must still
receive full up-front funding to begin new capital projects or acquire new
assets. And as noted above, some agency officials stated that the annual
mortgage payments may be even more of a dilemma than the up-front funding
requirement. Since a CAF assumes up-front funding, some agencies may still
seek to use some of the alternative financing mechanisms that they already
use, such as operating leases or enhanced-use leases, to meet capital
needs without first having to secure sufficient appropriations to cover
the full cost of the asset.41 As currently envisioned, CAFs would probably
not help improve capital planning concerns, such as the need for improved
budgeting and management of asset life-cycle costs. According to the NRC
report,42 operation and maintenance costs are typically 60 to 85 percent
of the total life-cycle costs of a facility while design and construction
typically account for only 5 to 10 percent of those costs.43 For example,
agencies must properly determine the funds needed for increasing staff in
new and expanded facilities in order to avoid staffing shortages.

Almost everyone we spoke with agreed that CAFs sounded complicated and
many questioned whether the challenges in budgeting for capital that CAFs
were designed to address were great enough to warrant CAFs as a solution.
Congressional budget committee and appropriations subcommittee staff
agreed that CAFs might be beneficial in theory but were probably not worth
the additional budget complexity they would create. Budget committee staff
considered the proposed benefits of a CAF to be abstract and uncertain
coupled with a sizeable likelihood for repayment problems in the out
years. In addition, they saw no obvious dilemma prompting the need for
CAFs. While this capital financing approach may be appealing in theory
since it promotes strategic planning and broadened, forward-looking
perspectives, budget practitioners cautioned the adoption of an approach
involving such layers of complexity in the absence of a clearly stated,
agreed-upon problem that the new approach is expected to address. Further,
they saw a need for agencies to

41For a thorough discussion of alternative financing approaches, see
GAO-03-1011.

42NRC, Investments in Federal Facilities, 27.

43Land acquisition, programming, conceptual planning, renewal or
revitalization, and disposal account for the remaining 5 to 35 percent.

complete their implementation of capital asset management and cost
accounting systems, which can help achieve some of the same benefits that
CAFs were meant to achieve. A good asset management system including
inventories and asset condition would likely be a necessary precursor to
successfully implementing CAFs. All of these factors weaken the case for
CAFs as an improved approach to current capital financing practices.

  Several Issues to Weigh When Considering Implementation of CAFs

While in theory CAFs could be implemented at most agencies, there are
several complex issues that Congress would need to consider before
adopting such a mechanism. For example, proposals to apply CAFs to
existing capital would require the development of a formula to calculate
an annual capital usage charge, which is likely to be a difficult and
contentious undertaking. Key players including OMB, CBO, and Congress
would need to work together to develop an agreed-upon method to estimate
an appropriate capital usage charge for various types of assets. And even
if the full cost of programs, including the cost of existing capital, was
more accurately reflected in the budget through the use of CAFs,
incentives to cut capital costs may not materialize except in times of
severe budget cuts. Even then, managers' abilities to eliminate unneeded
capital assets would probably be limited given mission responsibilities
and legal requirements that dictate the disposal of surplus federal
property. To remedy this, additional funding or agency flexibilities would
be needed, as would provisions to ensure debt repayment if CAF-financed
assets were transferred or sold. Additionally, it is likely that some
capital projects for federal office space, IT, and land would continue to
be financed outside of the CAF through mechanisms such as the FBF, WCFs,
or the GSA IT Fund.

    Imputing an Annual Capital Charge on Existing Capital May Offer Benefits but
    Would Be Difficult and Contentious

There are arguments that the CAF concept be applied to existing capital
assets as well as new capital assets to ensure that the full costs of all
programs are reflected in the budget. OMB points out that if CAFs were not
applied to all capital, it would be many decades before programs reflected
full annual costs and before the cost of alternative inputs could be
compared. Developing an annual capital usage charge for existing assets
would establish a level playing field for federal capital investment and
allow for comparisons across programs. In addition, this new charge could
influence agency managers to get rid of excess capital assets.

Accomplishing these goals would require developing a standard method of
computing an appropriate annual capital usage charge. Subunits would pay

these charges to the department's CAF using appropriated funds, which
would then be transferred to Treasury's general fund. In other words,
agencies would receive appropriations to pay for the use of capital assets
they already own and would not retain any of the funds to maintain or
replace assets. Imputing such a charge on existing capital is likely to be
difficult and very contentious given questions about how to estimate the
charge and the fact that the assets were already funded.

Before imputing an annual capital usage charge, key players, including OMB
and Congress, would need to agree on some type of standard formula to
estimate the charge. Three possible approaches to compute annual capital
usage charges would be to (1) use historical cost for the asset by
applying a charge as though the original cost had been financed by
borrowing from Treasury, (2) use market rental rates, or (3) devise a
calculation incorporating asset replacement cost, depreciation rates, and
interest rates. There are arguments for and against each of these options.
For example, while using historical cost would make the charge congruent
with accounting data; the charge would not reflect the current cost of
using capital and so might be less meaningful for evaluating costs.
Although using market rates would theoretically be the right measure for
comparing the cost of using resources for federal versus private purposes,
the fact that many government assets fill unique purposes means there is
not a measure of market value for them. For example, some agencies occupy
historic buildings, such as the Old Executive Office Building, for which a
comparable market-based value would be difficult to determine. The third
approach might be considered an agreeable middle ground, but applying
depreciation rates poses problems since they are largely arbitrary.
Agreement on whether to apply Treasury or market interest rates would be
necessary.

Some agency officials and congressional staff suggested that any charges
on existing capital should reflect the life-cycle costs of maintaining
assets and, similar to a WCF, receipts collected should be made available
for future maintenance and renovation costs. We have reported that repair
and maintenance backlogs in federal facilities are significant and that
the challenges of addressing facility deterioration are prevalent at major
real property-holding agencies.44 However, research and discussions on CAF

44GAO-03-122, 15.

design indicate that CAF receipts could only go to Treasury and not for
future projects. Officials were also skeptical about how to accurately
charge for highly specialized capital. For example, ARS has more than 100
laboratories located in various regions of the country, as well as abroad,
which are designed to carry out mission responsibilities ranging from the
study of crop production to human nutrition to animal disease control. The
highly technical and diverse nature of its objectives requires capital
assets that are suitable for varied climates, soils, and other
agricultural factors, which pose unique and difficult challenges in
establishing capital usage charges that would be viewed as acceptable by
agency officials.

If key players were able to agree on the method for calculating usage
charges on existing capital assets, they would also have to examine the
budgetary effects of such charges. Budget scorekeepers-OMB, CBO, and the
budget committees-would need to develop additional scoring rules to
clarify how the usage charges would be treated in the budget. Unlike
charges on new capital, there is no corresponding debt to repay. As a
result scorekeepers would have to specify how to score the usage charges
as they are transferred from the CAF to Treasury. Although these charges
would not change agency or government outlays or the deficit, they could
require a permanent increase in agencies' total BA, which would require
Congress to consider adjustments of appropriations subcommittee
allocations.45 Oversight would be especially important for these
transactions since CAF collections would be greater than needed to repay
Treasury loans, creating a temptation to use accruing balances for other
purposes.

Similar questions about how to charge for and how to score capital usage
charges for existing assets would eventually pertain to new capital funded
through the CAF. Once an asset is fully "paid off" through the CAF, it is
comparable to existing capital and would similarly incur an annual capital
usage charge. Some might argue that payments should continue in the same
amounts as before, while others may call for the calculation of a new
capital usage charge for "paid off" assets based upon the formula used for
capital that existed before the creation of CAFs. In any case, numerous
decisions on capital usage charges for existing capital would need to be
made prior to implementing CAFs.

45Although the scoring has not been determined, the subunit's payment to
the CAF for existing assets would likely require budget authority since
the payment is not for debt repayment.

Aside from the specifics of how to develop appropriate capital usage
charges, most agency officials and congressional staff with whom we spoke
were skeptical of the need for such a charge. Many said that the cost of
maintaining capital assets-which is reflected in agency budgets-and
depreciation expenses-which are reflected in agency accounting systems
along with asset maintenance costs-sufficiently represent the cost of
existing capital assets and help inform managers.46 As discussed earlier,
asset management systems and full cost accounting approaches are also
beginning to provide the information managers need to make better
decisions about the maintenance or disposal of existing assets and the
need for new capital. Some congressional staff thought the mind shift
required for Congress to agree to impute this new charge on existing
capital assets would be even more difficult than that required for
purchasing new capital using borrowing authority.

In the countries of New Zealand, Australia, and the United Kingdom,
charges on existing capital are being used to encourage the efficient use
of assets. These charges, similar to interest charges, are generally used
to reflect the opportunity cost of capital invested. In New Zealand,
departments are appropriated a capital charge based on their asset base at
the beginning of the year; at the end of the year they must pay the
government a capital charge based on their year-end asset base. If a
department has a smaller asset base at the end of the year than the asset
base for which the appropriation was made, the department is permitted to
keep part of the appropriation made for the capital charge. This spurred
the New Zealand Department of Education to sell a number of vacant sites
that it had acquired in the 1960s but that were no longer needed. However,
officials in New Zealand's Office of Controller and Auditor General were
uncertain about the effectiveness of having a charge for capital in
changing behavior significantly. In addition, some analysts in New Zealand
expressed concern that capital charging could drive department executives
to decisions that are rational in the short term but damaging in the long
term. For example, an audit official suggested that a department might
have an incentive to try to operate with obsolete and fully depreciated
assets in order to avoid a higher capital charge.

46OMB noted that neither the budget nor accounting systems reflect imputed
interest costs and therefore do not reflect full economic costs.

    Cost Allocation Efforts May Have Limited Effect on Agency Decision Making

Although one goal of CAFs is to ensure the allocation of full costs to
programs in the budget and thereby encourage managers to make more
informed decisions about capital assets, additional incentives to evaluate
new or existing asset needs are unlikely to be created except during times
of severe budget cuts or downsizing. For new assets funded through the
CAF, the mortgage payments made out of the subunits appropriations would
be equal to those received by the CAF and thus the payments would offset
each other within the department budget and at the appropriations
subcommittee level and would not affect the deficit. Although the
information on total program costs might be made more transparent, it is
not clear that this would create stronger incentives for more careful
deliberation on future asset needs than having these costs shown through
available methods such as cost accounting systems or the use of working
capital funds.

A charge on existing assets might also have limited impact. If
appropriation subcommittee allocations were simply raised to accommodate
new capital usage charges, programs would appear more expensive but
perhaps not differentially so. As with new assets, the capital charge on
existing assets would not affect the deficit. As a result, incentives for
rationalizing existing capital would not necessarily be created. Even
during tight budget years, when mandatory CAF payments would squeeze
operating budgets and be most likely to force trade-offs among capital
assets, managers may be constrained by mission responsibilities, legal
requirements, or the cost of disposing of assets. Consequently, agencies
might have to argue for increased funding or case-by-case exemptions,
which Congress has granted in the past.47

Some agencies questioned the effectiveness of applying a charge to
influence managers' decision making given the unique locations or types of
assets required to accomplish mission goals. BLM officials said an annual
capital usage charge would have a limited impact on their ability to
dispose of capital assets because of its stewardship role over the
nation's public lands. Similarly, ARS officials justified having locations
dispersed all over the country because its research activities are diverse
and require facilities in various climates and environments. As discussed,
Congress also plays a role in determining where ARS will conduct its
research. Likewise, many of

47For example, Congress has provided FS and BLM with specific authorities
to sell land in Los Angeles, California and near Las Vegas, Nevada. These
authorities allowed the agencies to retain the sale proceeds and use them
for new projects.

APHIS's capital assets are mission specific, including animal quarantine
stations, sterile insect-rearing facilities, and laboratories, and
typically do not have a comparable counterpart in the commercial sector.
APHIS officials said this limits managers' abilities to sell or transfer
assets because the land often must be converted to original condition, a
costly undertaking. For some subunits we spoke with, destruction of
certain assets, which also has an up-front cost, is the only viable option
for eliminating unneeded assets. For example, NPS and FS have many
facilities located on public land. If no longer needed, some of these
facilities cannot be sold or transferred and would have to be demolished.
According to FS officials, when they determine that an asset has exhausted
its useful life and needs to be disposed of, the agency will incur the
cost for removal and recover the salvage value.

Many agencies are subject to certain legal requirements that create
disincentives for disposing of surplus property. In these cases, agencies
would need additional funding or more flexibility to modify asset holdings
if improved decision making were to be realized. For example, under the
National Environmental Policy Act, agencies may need to assess the
environmental impact of their decisions to dispose of property. In
general, agencies are responsible for environmental cleanup of properties
contaminated with hazardous substances prior to disposal, which can
involve years of study and amount to considerable costs. Agencies that own
properties with historic designations-which is common in the federal
portfolio and certainly within the inventories of USDA and DOI-are
required under the National Historic Preservation Act to ensure that
historic preservation is factored into how the property is eventually
used. The Stewart B. McKinney Homeless Assistance Act, as amended, sets
forth a requirement that consideration be given to making surplus federal
property, including buildings and land, available for use by states, local
governments, and nonprofit agencies to assist homeless people.

If none of these restrictions apply and an agency is able to sell an
asset, most cannot retain the proceeds from the sale of unneeded property
even up to the cost of disposal. However, Congress has granted special
authorities in some cases. For example, FS officials told us it owned a
number of trails and roads on public lands that ran through the city of
Los Angeles, California. When the city expanded, it was no longer feasible
to maintain the roads and trails. As a result, the agency was granted
authority

to sell the land and use the proceeds to build a new ranger station.48 We
have said that agencies be allowed to retain enough of the proceeds from
an asset sale to recoup the cost of disposal, and that in some cases it
may make sense to permit agencies to retain additional proceeds for
reinvestment in real property where a need exists.49

    Issues Regarding Property Sales Would Further Complicate CAF Implementation

Provisions would also need to be established to ensure the full repayment
of CAF debts in the event that an agency sells or transfers a capital
asset before it reaches the end of its useful life (the repayment period).
Two possible options would be to (1) transfer the outstanding debt to a
new "owner" agency of the asset or (2) allow the "seller" agency to sell
the asset and use the proceeds from the sale to repay the outstanding CAF
debt. Both of these options would produce complications and issues to
resolve. For example, transferring the asset would require all parties
involved, including Treasury, to record adjustments to their CAF
accounting systems and oblige subunits to adjust their budget requests
accordingly. After the transfer, it is not clear whether the "seller"
agency's budget would be reduced by an amount equal to the asset's
mortgage payment. However, if that was done, it would lessen or eliminate
the incentive for the "seller" agency to sell or transfer the asset. If
the asset was sold instead of transferred, an appropriate "sale price"
would need to be determined as well as the appropriate disposition of the
sale proceeds. For example, if the asset was sold for an amount that is
greater than the outstanding CAF debt, the Treasury general fund would
receive full repayment on the asset plus excess revenue. On the other
hand, if an asset was sold for an amount less than the outstanding debt,
the CAF would default on the loan unless additional receipts for debt
repayment were appropriated. Finally, some subunits may argue to refinance
their mortgage if a lower Treasury interest rate became available and
lower payments would result. Again, before CAFs are implemented, proposals
on how to handle such circumstances would need to be addressed.

48In fiscal year 2002, Congress granted FS additional authorities to sell
or exchange excess buildings and other structures located on National
Forest System lands and under the jurisdiction of FS. The agency was
allowed to retain proceeds from the sales until the proceeds were expended
for maintenance and rehabilitation activities within the FS region in
which the building or structure was located. In fiscal year 2003, Congress
extended this authority allowing for some of the sale proceeds to be used
for construction of replacement facilities. Pub. L. No. 107-63, Sec. 329,
115 STAT. 471 (Nov. 5, 2001) and Pub. L. No. 108-7, Sec. 325, 117 STAT.
275-276 (Feb. 20, 2003).

49GAO-03-122, 41.

    Some Capital Would Likely Continue to Be Obtained through Existing Means

The CAF's scope of coverage would need to be addressed by any CAF
proposal. Capital assets are generally defined as land, structures,
equipment and intellectual property (such as software) that are used by
the federal government and have estimated useful lives of 2 years or more.
However, departments have some discretion in defining capital. The
Commission report suggested that OMB issue guidance on which capital items
belong in the CAF to ensure uniform implementation of the CAF proposal.
Alternatively, each department could use its current department guidelines
and definitions to determine which capital to fund through the CAF.
Whatever parameters are put in place, some capital assets would likely
continue to be funded outside the CAF through existing mechanisms.

For example, for federal office space, the Commission and NRC reports
state that agencies would generally continue to lease space from GSA and
pay rent to FBF. FBF, a governmentwide revolving fund, is used to acquire
office buildings and the space is then rented out to federal agencies.
Most agencies are not allowed to lease their own office space unless GSA
delegates its authority to do so to that agency, which GSA has done in the
past. Under the CAF mechanism, if GSA were to delegate this authority, the
CAF would lease the office space. The NRC report recommends that agencies
should use their CAF for office space acquisition only if it could be done
more effectively and efficiently than through GSA. GSA would negotiate the
acquisition of space for multiple agencies that seek to collocate in a
single facility.

Agencies also have the option to purchase IT through FTS and its IT Fund.
For a fee, FTS provides expertise and assistance in acquiring and managing
IT products. Those agencies that chose to use this service may argue for
continuing to finance these projects outside of the CAF so that they are
not paying a fee to FTS as well as interest on the borrowed funds. Some
officials also questioned the effectiveness of using borrowing authority
to finance IT purchases when their useful life is typically no more than
10 years and is often 5 years or less, thus indicating that officials may
argue to fund some IT projects outside the CAF. Departments and subunits
would also likely continue to rent certain capital assets from WCFs or to
use their WCFs to purchase some capital. As discussed, WCFs rely on user
charges to fund ongoing maintenance and replacement of capital assets and
the collections are used by some departments and subunits to finance
capital assets, such as vehicles and IT.

Land, such as wilderness areas, is also likely to remain outside the CAF.
Land retains its value so concepts such as depreciation and amortization
do

not apply to it. However, one subunit official stated that using borrowing
authority to buy land might be beneficial if it meant that land could be
purchased at a faster rate to obtain environmentally sensitive land before
it is damaged.

Conclusion	There is little doubt that in the mechanical sense CAFs could
work as a new system for financing capital assets. However, the
implementation and operation of the CAF concept would be complicated.
Managing the extra layer of responsibilities for CAF administration and
oversight would require the devotion of resources within departments,
subunits, and Treasury and to a lesser extent, OMB, CBO, and Congress.
Accounting for CAF transactions would be complex and burdensome. The
annual debt repayment would be a source of concern for Treasury and agency
officials, especially as more assets were financed through the CAF and
mortgage payments became a larger percentage of agency appropriations.

Beyond the complexities inherent in financing capital assets using
borrowing authority is a list of difficult issues that would have to be
resolved before benefits could be realized. The most difficult of these
issues, applying a capital usage charge to existing capital, would also be
the most important to address if annual capital costs were to be allocated
to program budgets. If CAFs were applied only to new assets going forward,
programs would not reflect the full annual cost of capital for decades and
programs purchasing new capital would appear more expensive than those
using existing capital. Even if this and other issues were tackled and
improved information about capital costs was provided to managers, there
is little assurance that CAFs alone would create incentives for programs
to reassess their use of capital. Even in times of severe budget
constraints, it is probable that managerial flexibility to adjust the
amount of assets used by a program would continue to be limited by agency
missions, legal restrictions, and limited funds for asset disposal. Given
the execution complexities and implementation concerns, the ensuing
question seems to be whether there are simpler methods that can be used to
achieve the same benefits as CAFs.

We believe there is strong evidence that both benefits attributed to CAFs
could be more easily obtained through existing mechanisms. Asset
management and cost accounting systems, when fully implemented, will be
important tools for promoting more effective planning and budgeting for
capital. Cost accounting systems can provide the same information on
capital costs as CAFs are intended to provide, while the information

provided by asset management systems could be even more crucial for
helping managers with limited budgets prioritize capital asset maintenance
and replacement. For existing capital, incentives to rationalize assets
might be created if agencies were allowed to retain proceeds to recoup the
cost of disposal, or in some cases, for reinvestment in real property.
While some of our case study agencies did not view spikes as a problem,
those that did felt they were managing them well through the use of WCFs,
no-year authority, and acquiring assets through useful segments. In any
case, spikes in spending for capital assets are likely to continue as
congressional and presidential priorities change over time.

When described in detail to executive branch and congressional officials,
we learned that the CAF proposal would likely have few proponents. Almost
everyone we consulted concluded that implementation issues would overwhelm
the potential benefits of a CAF. More importantly, current efforts under
way in agencies would achieve the same goals as a CAF without introducing
the difficulties. Given this, as long as alternative efforts uphold the
principle of up-front funding, then a CAF mechanism does not seem to be
worth the complexity and implementation challenges that it would create.

  Agency Comments and Our Response

We obtained comments on a draft of this report from OMB, Treasury, GSA and
our case study agencies-USDA and DOI. Treasury, GSA, USDA and DOI
generally agreed with the report. Treasury, USDA, DOI and OMB provided
technical comments, which have been incorporated as appropriate. OMB
agreed with our description of the mechanics of CAFs and concurred that
spikes in BA for capital assets could be alleviated through other means.
OMB also acknowledged the problems with CAFs that are highlighted in this
report, including those related to existing capital, and agreed that the
complications of designing and operating CAFs might outweigh the benefits.
However, they disagreed with our description of the primary goal of CAFs
and therefore do not believe alternative mechanisms achieve the same goal.

OMB supports having program budgets reflect full annual budgetary costs in
order to change incentives for decision makers. In addition to proposing
to budget for accruing retirement benefit costs, OMB has suggested
budgeting for accruing hazardous waste clean-up costs and budgeting for
capital through CAFs. Budgeting for full annual budgetary costs should
facilitate decision makers' ability to compare total resources used with
results achieved across government programs. For capital, OMB has

suggested CAFs as a possible method to allocate and embed the cost of
capital assets at the program budget level. OMB recognizes the usefulness
of asset management and cost accounting systems regardless of whether CAFs
are adopted. It is OMB's opinion that these tools do not ensure that the
costs of capital are captured in individual program budgets and therefore
do not affect incentives for decision makers in allocating resources among
and within programs. We disagree on several points.

We recognize that if the sole or primary purpose of a CAF is to embed
costs in the program budgets, then the alternatives discussed in this
report do not achieve that purpose. However we believe, as highlighted in
the Report of the President's Commission to Study Capital Budgeting, that
the primary goal of CAFs is to improve decision making for capital. We are
not convinced that CAFs and the annual mortgage payments they would
require would achieve this more effectively than other mechanisms. We
argue instead that the information provided by asset management and cost
accounting systems, when fully implemented, could assist decision makers
in efficiently allocating budgetary resources. While this information may
not necessarily be reflected in program budgets, it is available to aid in
budget and program decision making. The fact that many of these systems
are in relatively early stages of development also increases our concern
about CAFs. In a recent report, we noted the belief among some agency
officials, congressional appropriations committee staff, and budget
experts that improving underlying financial and performance information
should be a prerequisite to efforts to restructure program budgets.50 We
argue this would also be true for CAFs, since without adequate measures of
program costs and an ability to identify capital priorities, a new
financing mechanism would do nothing to address the basic challenges of
determining how much and what types of capital are needed.

It is also unclear that CAFs would create new incentives as OMB argues. As
we describe in the section titled "Cost Allocation Efforts May Have
Limited Effect on Agency Decision Making," if the annual mortgage payments
offset each other within the department budget and at the appropriations
subcommittee level, the deficit would not be affected, and it is unlikely
incentives would be changed. Even during tight budget years, when CAF
payments would squeeze operating budgets, managers may be unable to change
the amount of capital assets they use because of mission responsibilities,
legal requirements, or the cost of disposing of assets.

50GAO-05-117SP, 15.

We also recognize the value of linking resources to results in comparing
programs; however, it is unclear that CAFs are necessary or would even
work to accomplish this. Institutionalizing CAFs could permit program
comparison, but fair evaluations would only be possible if existing
capital were included. Therefore, the difficult issue of including
existing capital would have to be addressed. Alternatively, we believe
that cost accounting systems, when well developed within and across
agencies, provide a similar opportunity for comparing programs. In
conclusion, we remain of the view that the operational challenges of CAFs
outweigh the benefits and that alternative mechanisms described in this
report can more simply accomplish the goals of CAFs.

As we agreed with your office, unless you publicly announce the contents
of this report earlier, we plan no further distribution of it until 30
days from its issuance date. At that time we will send copies of this
report to the Director of the Office of Management and Budget, the
Administrator of the General Services Administration, the Secretary of the
Department of the Interior, the Secretary of the Department of
Agriculture, and the Secretary of the Department of the Treasury. We will
also make copies available to others upon request. This report will also
be available at no charge on the GAO Web site at http://www.gao.gov. If
you or your staff have any questions regarding the information in this
report, please contact me at (202) 512-9142 or Christine Bonham at (202)
512-9576. Key contributors to this report were Jennifer A. Ashford, Leah
Q. Nash, and Seema V. Dargar.

Sincerely yours,

Susan J. Irving Director, Federal Budget Analysis Strategic Issues

Appendix I

Comments from the Department of the Treasury

Note: GAO comments supplementing those in the report text appear at the
end of this appendix.

Appendix I
Comments from the Department of the
Treasury

                                   Appendix I
                      Comments from the Department of the
                                    Treasury

GAO's Comments	We believe that the discussion of BPA's use of "fish
credits" is an appropriate example for the section on agencies' repayment
of their borrowing from Treasury. Although these credits were provided by
Congress, their use for offsetting payments on Treasury debt has been
controversial and opposed by some members of Congress and other interested
parties. However, we have made technical changes to the section based on
Treasury's comments.

Appendix II

Comments from the Department of the Interior

Appendix II
Comments from the Department of the
Interior

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

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

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

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

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

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

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

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

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

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

Relations Washington, D.C. 20548

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

                               Presorted Standard
                              Postage & Fees Paid
                                      GAO
                                Permit No. GI00

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

Official Business
Penalty for Private Use $300

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