Weapons Acquisition: Warranty Law Should Be Repealed (Chapter Report,
06/28/96, GAO/NSIAD-96-88).

GAO reviewed the Department of Defense's (DOD) use of major weapon
system warranties, focusing on whether these warranties: (1) provide
expected benefits to the government; and (2) are compatible with the
weapon systems acquisition process.

GAO found that: (1) DOD receives about $1 in direct benefit for every
$19 paid to a contractor for a warranty; (2) DOD program officials
rarely seek to waive the warranty requirement because waivers require
the approval of the Secretary of Defense, an Assistant Secretary of
Defense, and congressional defense authorization and appropriations
committees; (3) despite DOD regulations that require a cost-benefit
analysis to determine if the warranty is cost-effective, some
cost-benefit analyses are inadequate; (4) the military services are not
conducting post-award assessments to determine whether warranty costs
are commensurate with the benefits received and to identify advantageous
and disadvantageous warranty provisions for future contracts; (5) while
warranties may provide insurance against catastrophic financial loss or
excessive repair or replacement costs, assurance that the products
conform to the design, quality, and performance levels specified in the
contract, and incentives that motivate contractors to improve product
quality, none of these benefits justify the expense of buying a
warranty; and (6) warranties only extend the period that DOD can
determine that a product does not conform to contract specifications and
requirements and require the contractor to make repairs.

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

 REPORTNUM:  NSIAD-96-88
     TITLE:  Weapons Acquisition: Warranty Law Should Be Repealed
      DATE:  06/28/96
   SUBJECT:  Advanced weapons systems
             Weapons industry
             Warranties
             Cost effectiveness analysis
             Defense cost control
             Maintenance costs
             Repair costs
             Defense procurement
             Department of Defense contractors
             Federal procurement policies
IDENTIFIER:  F-15E Aircraft
             F-16 Aircraft
             Multiple Launch Rocket System
             MLRS
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


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


Report to Congressional Committees

June 1996

WEAPONS ACQUISITION - WARRANTY LAW
SHOULD BE REPEALED

GAO/NSIAD-96-88

Weapons Acquisition

(707099)


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

  FAR - Federal Acquisition Regulation
  DFARS - Defense Federal Acquisition Regulation Supplement
  DOD - Department of Defense
  GAO - General Accounting Office

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


B-259383

June 28, 1996

Congressional Committees

This report addresses the use of major weapon system warranties
required by 10 U.S.C.  2403 and discusses whether weapon system
warranties (1) provide the expected benefits to the government and
(2) are compatible with the weapon systems acquisition process.  This
report contains recommendations to the Secretary of Defense and a
recommendation to the Congress, both of which are intended to improve
weapon system warranty policies and processes.  We conducted this
review under our basic legislative responsibilities. 

We are sending this report to you because of your Committees'
interest in these issues.  Copies are also being sent to the
President of the Senate; Speaker of the House of Representatives;
Secretaries of Defense, the Air Force, the Army, and the Navy; and
the Director of the Office of Management and Budget.  We will also
make copies available to others upon request. 

This report was prepared under the direction of Louis J.  Rodrigues,
Director, Defense Acquisitions Issues, who may be reached on (202)
512-4841 if you or your staff have any questions concerning this
report.  Other major contributors are listed in appendix III. 

Henry L.  Hinton, Jr.
Assistant Comptroller General


List of Congressional Committees

The Honorable Strom Thurmond
Chairman
The Honorable Sam Nunn
Ranking Minority Member
Committee on Armed Services
United States Senate

The Honorable Floyd Spence
Chairman
The Honorable Ronald Dellums
Ranking Minority Member
Committee on National Security
House of Representatives



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


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

As part of the 1985 Department of Defense (DOD) Authorization Act,
Congress enacted legislation requiring DOD to obtain cost-effective
warranties on weapon systems.  The warranties were expected to
improve weapon system reliability by providing a mechanism to hold
contractors liable for poor performance.  Past reviews by GAO\1 and
others concluded that DOD was not properly managing its warranty
program.  This report assesses whether the warranties being obtained
for weapon systems provide the expected benefits to the government
and whether warranties, as required by 10 U.S.C.  2403, are
compatible with weapon system acquisitions. 


--------------------
\1 DOD Warranties:  Improvements Needed in Implementation of Warranty
Legislation (GAO/NSIAD-87-122, July 21, 1987) and DOD Warranties: 
Effective Administration Systems Are Needed to Implement Warranties
(GAO/NSIAD-89-57, Sept.  27, 1989). 


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

The warranty provisions contained in the 1985 DOD Authorization Act
and codified as 10 U.S.C.  2403 changed DOD's approach to
administering weapon system warranties.  The legislation requires
warranties on weapon systems that have a unit cost of more than
$100,000 or an expected total procurement cost of more than $10
million.  The law requires the prime contractor to guarantee that the
item will (1) conform to the design and manufacturing requirements
delineated in the contract, (2) be free from all defects in materials
and workmanship at the time it is delivered to the government, and
(3) comply with the essential performance requirements delineated in
the contract. 

DOD, in 1992, proposed repealing the warranty law.  In 1993, the
Acquisition Law Advisory Panel also recommended that Congress repeal
the warranty law, citing the heavy administrative burden the law
imposes and DOD's failure to realize expected warranty benefits.  The
Federal Acquisition Streamlining Act of 1994 (P.  L.  103-355)
deleted the requirement for an annual waiver report to Congress, but
continued to require waiver notifications to the defense committees. 
It also directed the Office of the Secretary of Defense to establish
guidelines for obtaining cost-effective warranties and seeking
waivers when necessary. 


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

Requiring the use of warranties in weapon system acquisitions is not
practical and does not provide the government much in the way of
benefits.  GAO estimated that the military services spend about $271
million annually for weapon system warranties.  These expenditures
have resulted in a financial return of approximately 5 cents for
every dollar spent.  Congress expected that warranties would ensure
that contractors provide quality weapon systems to DOD or be held
accountable and would provide an incentive for contractors to improve
the weapon systems.  However, having DOD obtain warranties, in
practice, is an expensive method of resolving product failures with
the contractor. 

Although Congress never intended for DOD to obtain warranties that
were not cost-effective, and provided authority in the law for
waivers, program officials are discouraged from seeking waivers by a
statutory waiver process which has become burdensome and protracted. 
Consequently, warranties are almost always obtained.  Further, the
services are not generally conducting post-award assessments, which
could provide the basis for deciding whether or not to obtain future
warranties. 

Even though DOD is moving toward a more commercial acquisition
system, it retains many methods to validate the quality of weapon
systems.  It is often involved in designing the system and approving
the production and quality assurance processes used by the
manufacturer.  It is these methods that, while not always successful,
focus on developing and producing a high quality weapon system.  A
warranty simply provides a way to obtain repairs or monetary
compensation from the contractor, after a problem has occurred,
without necessarily contributing to the overall quality of the
product. 

The government has traditionally self-insured because its large
resources make protection against catastrophic loss unnecessary. 
Further, it is often the only buyer for a product and cannot share
the insurance cost with other buyers.  As a result, a contractor
cannot allocate the cost of insuring against the risk of failure
among multiple buyers.  DOD will bear the entire estimated cost. 
Also, DOD program officials told GAO that warranties do not motivate
contractors to improve the quality of their products.  As a result,
requiring the use of warranties in weapon system acquisition is not
practical and does not provide the government much in the way of
benefits.  GAO believes the warranty law should be repealed and the
decision to obtain a warranty, in appropriate cases, should be left
to the program manager. 


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


      DOD'S WARRANTY COSTS EXCEED
      BENEFITS
-------------------------------------------------------- Chapter 0:4.1

Based on the contracts GAO reviewed, DOD's costs for warranties have
greatly exceeded any financial return it has received.  For contracts
on which DOD could provide both price and claim data, GAO estimated
that DOD received about $1 in direct benefit for every $19 paid to a
contractor for a warranty.  One cause of this negative return is the
low submission rate for warranty claims. 

Conceptually, warranties provide unquantifiable benefits such as
prepaid maintenance support and a mechanism for resolving product
performance disputes.  For example, under a warranty, a contractor
performs repairs and charges them against the warranty.  In the end,
however, the contractor profits by keeping the difference between
actual claims and the warranty price, a total of $89 million for the
eight contracts GAO reviewed that had both price and claims data.  In
addition, although some military service officials said that repairs
were performed quicker under a warranty, others said the reverse was
true and contractors routinely contested warranty claims. 


      WAIVER PROCESS IS NOT USED
      AS EXPECTED
-------------------------------------------------------- Chapter 0:4.2

Obtaining a waiver of the warranty requirement is not considered a
viable alternative by the majority of acquisition officials GAO
interviewed.  Under 10 U.S.C.  2403, only the Secretary of Defense or
an Assistant Secretary of Defense or a military service can waive the
warranty requirement if the warranty is (1) not cost-effective or (2)
not in the interest of national defense.  In addition, congressional
defense authorization and appropriations committees must be notified
before a waiver is granted for major weapon systems.\2 Both
requirements discouraged program officials from seeking waivers, even
when a cost-effectiveness analysis clearly showed that a warranty was
not in the best interest of the government.\3 Since 1985, only 21
waivers have been requested DOD-wide, and 15 were approved. 


--------------------
\2 For 10 U.S.C.  2403, a major weapon system is one subject to the
Selected Acquisition Reports in
10 U.S.C.  2432. 

\3 Prior to the Federal Acquisition Streamlining Act of 1994, DOD
also had to submit an annual warranty waiver report for waivers that
were not covered by the notification requirement. 


      COST-BENEFIT ANALYSIS ARE
      NOT ADEQUATE
-------------------------------------------------------- Chapter 0:4.3

DOD regulations require that a cost-benefit analysis be performed and
documented in the contract file to determine if the warranty is
cost-effective.  The Navy conducted only one cost-benefit analysis in
the contracts reviewed by GAO.  For the Air Force and Army contracts
we reviewed, GAO found the cost-benefit analyses that were conducted
were inadequate because (1) warranties were sometimes not separately
priced, thereby omitting a major cost; (2) the analyses failed to
account for the government's administrative costs; (3) the analyses
often assumed that all potential defects would be identified and
claims submitted; and (4) the services did not conduct present value
analyses, which allow the comparison of current expenses with future
benefits. 


      WARRANTY POST-AWARD
      ASSESSMENTS ARE GENERALLY
      NOT BEING PERFORMED
-------------------------------------------------------- Chapter 0:4.4

Post-award assessments are required by each of the services'
regulations.  These include an in-process assessment required by all
the services, and a final payoff assessment only required by the
Army.  These assessments are used to determine whether the warranty
costs are commensurate with the benefits received and to identify the
advantageous and disadvantageous warranty provisions for future
contracts.  The military services, however, conducted post-award
assessments for only 3 of the 38 contracts GAO reviewed. 


      WEAPON ACQUISITIONS SELDOM
      BENEFIT FROM WARRANTIES
-------------------------------------------------------- Chapter 0:4.5

Conceptually, a warranty in weapon system acquisitions may provide
three functions:  insurance, assurance-validation, and
incentivization.\4 These functions have distinct purposes; however,
none of them have really worked to DOD's advantage and GAO did not
find any cases that justified the expense of buying a warranty to
perform one of these functions. 

  -- The insurance function protects a buyer against catastrophic
     financial loss or excessive operating costs as well as the risks
     of repair or replacement costs.  This concept is flawed in
     weapon system acquisitions because insurance is based on the
     principle of shared risk.  However, the government is usually
     the only purchaser of a weapon system and the contractor cannot
     allocate the cost of insuring against that risk among multiple
     buyers.  The complete cost for the estimated risk must be borne
     by the sole buyer or absorbed by the contractor.  If it is borne
     by the buyer it becomes the price of the warranty.  If it is
     absorbed by the contractor, then it is a cost that must be
     covered by the price of the system.  In both cases the buyer,
     DOD, pays.  In addition, even a large, key defense contractor
     would most likely be unable to absorb a catastrophic loss and
     would be protected by the government from going out of business
     in attempting to pay that loss. 

  -- The assurance-validation function ensures that all the products
     conform to the design, quality, and performance levels specified
     in the contract.  However, DOD is extensively involved in
     creating the design, approving the production processes, and
     conducting tests and inspections of the weapon system.  The only
     potential benefit this warranty function appears to provide the
     government is protection against the failure of these processes. 

  -- The incentivization function theoretically motivates a
     contractor to improve product quality, thereby minimizing any
     repair costs.  However, the vast majority of procurement
     officials GAO interviewed do not believe they induce the
     contractors to improve the quality of weapon systems. 
     Contractors calculate the cost of projected failure rates and
     repair costs and include these costs in the warranty price. 

The only real benefit that a warranty seemed to provide is to extend
beyond acceptance the period in which DOD can determine that a
product does not conform to contract specifications and requirements
and require the contractor to make repairs. 


--------------------
\4 Robert E.  Kuenne, Paul H.  Richanbach, Frederick R.  Ridell, and
Rachel Kaganoff, Warranties in Weapon System Procurement:  Theory and
Practice, Westview Press, (1988). 


   RECOMMENDATIONS TO THE
   SECRETARY OF DEFENSE
---------------------------------------------------------- Chapter 0:5

GAO recommends that the Secretary of Defense establish an expedited
waiver process that will limit the disincentives inherent in the
current process.  GAO also recommends that the Secretary of Defense
revise DOD's acquisition policies to adequately manage those
warranties that the military services determine should be obtained. 
Consideration should be given to (1) requiring that all weapon system
warranties be separately priced in order to allow meaningful
cost-benefit analyses; (2) improving cost-benefit analyses through
more realistically reflecting the likelihood of claim submission,
performing present value analyses, and including the government's
administrative costs; and (3) ensuring that the services enforce the
regulations requiring post-award assessments of weapon system
warranties so that the services will know why these warranties were
or were not beneficial to the government.  GAO also recommends that
the Secretary of Defense direct the Secretaries of the Air Force and
the Navy revise their regulations to specifically require a final
payoff assessment for weapon system warranties as the basis for
purchasing more beneficial follow-on warranties and building
institutional knowledge for procuring and administering effective
warranties. 


   RECOMMENDATION TO CONGRESS
---------------------------------------------------------- Chapter 0:6

The administrative problems that GAO has identified appear to be
unintended consequences of the warranty law due to the de facto
mandatory nature of warranties.  Attempts to administratively correct
the problems have not been very successful.  Since DOD continues to
have problems administering weapon system warranties and the
warranties provide minimal benefits for the costs incurred, Congress
should
repeal 10 U.S.C.  2403.  Were the warranty requirement repealed, DOD
and the services would still have management flexibility to obtain
warranties for major weapon systems when deemed appropriate.  As was
done prior to the warranty law, DOD and the services would rely on
the Federal Acquisition Regulation and their own policies to
determine when it is appropriate to obtain a weapon system warranty. 
The decision should be documented as part of the system acquisition
strategy. 


   AGENCY COMMENTS AND GAO'S
   EVALUATION
---------------------------------------------------------- Chapter 0:7

In commenting on a draft of this report, DOD stated that it "strongly
supports" GAO's recommendation that Congress should repeal 10 U.S.C. 
2403.  Since 1992, DOD has supported the need for congressional
repeal of the weapon system warranty law.  DOD only partially
concurred with the recommendations to the Secretary of Defense,
stating that the solution to the problems cited by GAO is repeal of
the law.  DOD indicated that it will ask the military departments to
review their warranty waiver process.  DOD noted, however, that in
order to remove the disincentives and streamline the waiver process
it needs relief from the congressional notification requirement and
the warranty waiver approval level. 

DOD stated that it did not see the need to separately price
warranties because it has insight into warranty costs through cost
reporting and can project warranty cost from actual claim data. 
Although DOD currently has insight into warranty costs, GAO found
that this information is often not used in the cost-benefit analyses. 
Therefore, GAO believes that separately pricing the warranty would
permit DOD to perform better warranty cost-benefit analyses. 

DOD stated that all the military departments have in place
regulations that require post-award assessments, but acknowledged
that the military departments were not fully complying with existing
regulations.  DOD stated that it would reiterate the importance of
such assessments in a memorandum to the military departments.  GAO's
review indicated, however, that only the Army's regulation
specifically requires a final pay-off assessment to determine the
economic benefit derived from a warranty.  GAO believes that the Air
Force and the Navy regulations should be revised to explicitly
require final payoff assessments. 


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

In the 1970s and the 1980s, Congress received numerous reports about
problems with the weapon acquisition process, namely that weapon
systems often failed to meet their military missions, were
operationally unreliable, and had defects in materials or
workmanship.  To address manufacturing deficiencies and performance
shortcomings, Congress began requiring the Department of Defense
(DOD) to obtain written warranties on all production contracts for
weapon systems costing over $100,000 per unit or whose eventual
acquisition cost is more than $10,000,000.\1 Congress expected that
obtaining cost-effective warranties would enable DOD to hold
contractors accountable for the performance of their systems and that
the risk of financial consequences would encourage contractors to
improve the quality and reliability of the systems. 

In 1984, when the warranty provision was first enacted, many DOD and
industry officials criticized the law as being impractical,
unworkable, and potentially costly.  An amended version enacted in
the 1985 DOD Authorization Act and codified as 10 U.S.C.  2403, was
intended to correct the problems. 

For most non-weapon system purchases, the Federal Acquisition
Regulation (FAR) prescribes the procedures and purposes of obtaining
a warranty.  Under the FAR, the use of a warranty is not mandatory. 
The FAR allows contracting officers to require contractors to provide
warranties on products sold to the government.  The decision is based
on a determination that a warranty would be in the government's best
interest.  In addition, the Defense Federal Acquisition Regulation
Supplement (DFARS) provides additional guidance on when it is
appropriate to obtain a weapon system warranty. 


--------------------
\1 The statute uses the term guarantee.  However, warranty is the
more commonly used term to describe the extended contractual
relationships envisioned by the law, so we will use that term in this
report. 


   WARRANTY LAW REQUIREMENTS
---------------------------------------------------------- Chapter 1:1

Under 10 U.S.C.  2403, an agency head is prohibited from entering
into a production contract for a weapon system with a per unit cost
greater than $100,000, or a total system cost over $10 million,
unless the prime contractor provides a warranty.  The prime
contractor must warrant that items provided under the contract (1)
conform to the design and manufacturing requirements delineated in
the contract, (2) are free from all defects in materials and
workmanship at the time of delivery, and (3) meet the essential
performance requirements delineated in the contract.\2 Contractors
are not required to provide a warranty on government-furnished
equipment. 

If the Secretary of Defense determines that a warranty is not in the
interest of national defense or that a warranty will not be
cost-effective, he may waive all or part of the warranty requirement. 
The Secretary cannot delegate the waiver authority below the level of
an Assistant Secretary of Defense or of a military department.  The
Secretary must also notify the Senate Committee on Armed Services and
the House Committee on National Security before granting a waiver for
a major weapon system.\3


--------------------
\2 As defined in the statute, "essential performance requirements"
are the operating capabilities or maintenance and reliability
characteristics of the system that the Secretary of Defense
determines are necessary for the system to fulfill the military
requirement for which it is designed. 

\3 For 10 U.S.C.  2403, a major weapon system is one subject to the
Selected Acquisition Reports of
10 U.S.C.  2432. 


   TYPES OF WARRANTIES
---------------------------------------------------------- Chapter 1:2

Generally, warranties require that the contractor repair or replace
noncomplying or defective goods covered by the warranty without cost
to the government and/or pay the government's costs of correcting the
defective condition.  Warranted defects or deficiencies may be caused
by poor design, faulty manufacturing processes, or the use of
materials that do not meet contract specifications.  The cost and
coverage of warranties are negotiated on a contract-by-contract
basis.\4 Typical weapon system warranties fall into one of the
following three categories:  failure-free, threshold, and systemic. 

When a system is covered by a failure-free warranty, the contractor
is obligated to correct all defects that occur during the warranty
period.  Although a failure-free warranty is easy to implement, it is
associated with high costs due to the higher risks assumed by the
contractor.  A threshold warranty requires a contractor to remedy a
defect when a threshold, such as a predetermined number of part or
system failures, is exceeded.  This type of warranty recognizes that
all weapon systems malfunction to some degree, and the warranty only
requires action if the weapon system does not meet the agreed-upon
reliability levels. 

A systemic warranty covers a system against a defect that occurs with
regularity throughout a production lot or fleet.  In the case of
systemic warranties, the government must prove that the defects are
occurring regularly by either conducting its own investigation or
supervising an investigation by the contractor.  Once the government
proves that a systemic defect exists, the contractor is responsible
for replacing or repairing all of the items produced under the
circumstances that caused the defect.  Some systemic warranties also
require the contractor to redesign warranted items if the defect is
the result of a design problem. 

A DOD weapon system may be covered by multiple types of warranties. 
For example, an item may be covered by a failure-free warranty until
it is transferred to a unit, and then covered by a systemic warranty. 


--------------------
\4 Negotiable items include the duration of the warranty and the
maximum financial exposure of the contractor under the warranty. 


   PRIOR REVIEWS
---------------------------------------------------------- Chapter 1:3

In 1987, we reported\5 that the military services were obtaining
warranties without assessing cost-effectiveness.  We also found that
warranty terms and conditions were not clearly stated in most
contracts.  Also, many warranties did not delineate whether redesign
was a remedy if performance requirements were not met.  We concluded
that this situation could result in warranty administration problems. 

In 1989, we reported\6 that (1) the Office of the Secretary of
Defense was not actively overseeing warranty administration by the
services; (2) the services had not established a fully effective
warranty administration system; (3) the procurement activities had
problems performing cost-effectiveness analyses; and (4) the
services, therefore, did not know whether they should seek warranty
waivers.  We concluded that DOD had little assurance that warranty
benefits were being fully realized. 

DOD's Director for Defense Procurement, in 1992, proposed repealing
the warranty law.  This initiative was included as Section 620 of
DOD's Legislative Program for the 103rd Congress. 

In a January 1993 report,\7 DOD's Acquisition Law Advisory Panel,
referred to as the Section 800 Panel, recommended repealing the
warranty law based upon two reviews\8

that highlighted significant problems with the administration and
effectiveness of the law.  These reviews found that (1) waiver
requests were not seriously considered, (2) the use of waivers had
been "virtually nil," (3) contractor expenses for warranty repairs
were less than the negotiated price for the warranty in four out of
five cases, (4) only two out of seven threshold warranties ever
reached the threshold, (5) no claims had been made on systemic
warranties reviewed, and (6) service regulations requiring post-award
reviews of warranty cost-effectiveness were not enforced.  The
Panel's alternate recommendation was to revise 10 U.S.C.  2403 to
address the implementation problems.  The Section 800 Panel sought
greater flexibility in implementing and tailoring warranties, as well
as limiting warranties to major weapon systems.  Furthermore, the
Section 800 Panel recommended that the waiver approval authority be
lowered from the Assistant Secretary level and that a policy
statement be issued encouraging the use of waivers when a warranty is
not cost-effective. 

Congress did not repeal the warranty law.  Instead, the Federal
Acquisition Streamlining Act of 1994 (P.L.  103-355) modified the
congressional notification requirement so that an annual report of
waivers granted is no longer required, although the defense
committees are still to be notified before a waiver is granted for a
major weapon system.  The act also required DOD to issue guidance on
negotiating cost-effective warranties and on waivers.  In response,
DOD revised subpart 246.7 of DFARS to stress that the use of weapon
system warranties may not be appropriate in all situations and that a
waiver should be obtained if a warranty is not cost-effective or in
the interest of national defense. 


--------------------
\5 DOD Warranties:  Improvements Needed in Implementation of Warranty
Legislation (GAO/NSIAD-87-122, July 21, 1987). 

\6 DOD Warranties:  Effective Administration Systems Are Needed to
Implement Warranties (GAO/NSIAD-89-57, Sept.  27, 1989). 

\7 DOD's Acquisition Law Advisory Panel, Streamlining Defense
Acquisition Laws (Jan.  1993), pp.  2-113 to 2-119. 

\8 The reviews were MKI, Inc., Warranty Guidebook Research Summary,
Defense Systems Management College (1992) and Office of the Deputy
Director for Defense Systems Procurement Strategies, Report on the
Administration of Department of Defense Weapon System Warranties
(Sept.  1992). 


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

Our objectives were to determine whether the warranties being
obtained for weapon systems provide the expected benefits to the
government, and to assess whether the use of warranties, as required
by law, is compatible with the acquisition of weapon systems.  We
analyzed the warranty legislation, DOD and service policy guidance
and regulations, and procurement activity guidelines governing the
use of warranties in weapon system acquisitions.  To obtain insight
into the types of issues faced in managing a warranty program, we
gathered warranty information from 22 ongoing acquisition programs
and reviewed the results of warranty studies performed by the DOD
Inspector General, the Office of Defense Procurement, the Acquisition
Law Advisory Panel, and others.  We selected systems based on the
contract value and the type of weapon system for contracts awarded
between 1984 and 1994.  Our report focuses on the use warranties for
DOD major weapon systems and does not cover the use of warranties on
commercial subcomponents in weapon systems or commercial items.  In
some instances, the information available in contract files was
limited because the services had not collected the information or
there was a lack of centralized documentation. 

Our work was performed primarily at the six commands responsible for
managing the major acquisition programs we selected for our review. 
The following are the procurement commands visited: 

Air Force

Air Force Material Command

Army

Aviation and Troop Command
Missile Command
Tank-Automotive and Armaments Command

Navy

Naval Sea Systems Command
Naval Air Systems Command

At the procurement commands, we reviewed contract files, including
basic contract information, warranty and inspection clauses,
cost-effectiveness studies, and correspondence.  We supplemented the
information by interviewing program management, as well as defense
contracting, policy, and legal officials.  We also held discussions
with officials from the Office of the Secretary of Defense and the
Defense Systems Management College.  In addition, we contacted
selected contractor and professional association officials to obtain
their viewpoints on the advantages and disadvantages of using
warranties in major weapon system acquisitions. 

We performed our review from November 1994 through February 1996 in
accordance with generally accepted government auditing standards. 


WEAPON SYSTEM WARRANTIES ARE
EXPENSIVE AND HAVE PROVIDED LITTLE
BENEFIT
============================================================ Chapter 2

DOD is obtaining weapon system warranties that are not cost-effective
because it does not use waivers as expected by Congress and does not
perform adequate cost-benefit analyses or post-award assessments to
ensure that the decisions to obtain or not to obtain a warranty are
based on a valid foundation.  Congress did not intend for DOD to
obtain warranties that were not cost-effective.  Therefore, the
warranty law allows the Secretary of Defense to waive the use of a
warranty if the Secretary determines that it would not be
cost-effective.\1 However, none of the warranties we reviewed, where
claim and price data was available, were cost-effective.  We found
that the government paid $94 million and collected $5 million on
these weapon system warranties.  We also calculate that the military
services spend approximately $271 million annually to pay for
warranties.\2 Further, this cost is only the warranty price paid to
the contractor.  It does not include the additional costs to the
government of negotiating and administering warranties.  Reviews by
others have also found that weapon system warranties are generally
not cost-effective. 


--------------------
\1 See subsection (d) of 10 U.S.C.  2403.  A waiver may also be
granted if the Secretary determines that the waiver is necessary in
the interest of national defense. 

\2 Based on our case studies, we calculated that the average price of
a warranty, for the contracts we reviewed, is about 0.87 percent of
the total contract's value.  We applied this percentage to the
average amount of the contracts awarded by DOD for fiscal years
1992-95.  We obtained the contract values from DOD's DD350 database. 


   WEAPON SYSTEM WARRANTIES ARE
   NOT COST-EFFECTIVE
---------------------------------------------------------- Chapter 2:1

Warranties have both quantified and unquantified costs.  The
quantified cost is the negotiated price for the warranty, while the
unquantified cost includes the negotiation and subsequent
administration of warranties.  Warranties also provide both
quantified and unquantified benefits.  The quantified benefit to the
government includes financial compensation received as a result of
claims and low-cost or no-cost proposals to correct problems, while
the unquantified benefits claimed by program officials include
prepaid maintenance support for field units and the value of having a
process in place for readily resolving product performance problems. 

We found that the weapon system warranties purchased by DOD were not
cost-effective.  We were able to obtain warranty price and claim data
on four weapon systems and eight contracts.  In every case where
price and claim data was available, the warranty price exceeded the
value of the claims made.  The combined warranty price was $94
million, the value of the warranty claims was $5 million, and the
quantified price exceeded the quantified benefit by $89 million. 
(See app.  I.) For example: 

  -- The government paid $12 million for the F-15E (1) design and
     manufacture and (2) materials and workmanship warranties for the
     1989 and 1990 contracts, which covered the purchase of 72
     aircraft.  The program office identified 260 potential warranty
     claims, of which 134 were agreed to and corrected by the
     contractor.  There were 126 claims that were not agreed to by
     the contractor for a variety of reasons, including the fact that
     failed parts were unavailable for contractor inspection.  The
     program office estimated that the average cost to fix each
     problem was $3,000 and that the total financial benefit to the
     government was $402,000.  The quantified costs, therefore,
     exceeded the quantified benefits by about $11.6 million. 

  -- The F-16 Multiyear II warranty price for 720 aircraft procured
     between 1986 and 1989 was $27.86 million.  In a 1991 study,\3
     the program office calculated that the warranty benefit was
     $2.78 million, or about 10 percent of the warranty price.  While
     the warranty coverage had not expired at the time, the study did
     project a total potential benefit of $9.94 million for this
     warranty, or 36 percent of the warranty price.  The study found
     "little tangible return on investment" for this warranty.  The
     program office was unable to provide final claim figures. 

  -- The Multiple Launch Rocket System 1985 warranty cost $1.584
     million.  The estimated value of the warranty claims was
     $126,000.  Therefore, the quantified cost exceeded the
     quantified benefit by $1.458 million. 

In 1992,\4 the Army found a similarly large imbalance between the
costs incurred and the total dollars recovered under several
warranties.  A review of 36 expired warranties on 12 weapon systems
at the Missile Command through December 1990 showed that the warranty
cost for these contracts was $27.9 million and the dollar value of
the warranted repairs was $12.5 million--meaning that these
warranties had a negative monetary return on investment of $15.4
million. 

The Air Force has also recognized that it has been obtaining some
non-cost-effective warranties.  The Deputy Assistant Secretary of the
Air Force for Contracting, stated in a memorandum in 1992, ".  .  . 
we agree that warranties are not always cost-effective.  Recent
experience indicates that contractors are unwilling to provide
reasonable cost proposals in some cases, even when historical
warranty cost data is available that suggests a much lower warranty
price is appropriate."

The warranty price does not include all warranty costs to the
government.  The costs not included in the warranty price are
associated with warranty development, administration, training, the
need to obtain and provide special data, in-plant warranty
monitoring, special transportation, increased spare component
requirements because of longer logistical repair times, decreased
competition opportunities, and reduced self-sufficiency of the
military services.  We did not estimate these additional costs to the
government.  In some cases, we had no basis for an estimate and in
others the additional costs due to the warranty could not be readily
identified. 


--------------------
\3 Analysis of F-16 Multiyear II Engineering Change Proposals for
Correction of Deficiencies, F-16 System Program Office (Feb.  22,
1991). 

\4 Evaluation of the U.S.  Army Warranty Program, U.S.  Army Materiel
Systems Analysis Activity (Technical Report No.  547, Aug.  1992), p. 
49. 


      NOT ALL POTENTIAL CLAIMS
      SUBMITTED
-------------------------------------------------------- Chapter 2:1.1

One cause for the quantified cost exceeding quantified benefits is
the low claims submission rate for warranted items.  Air Force
officials told us that one reason for the low claims rate is that
submitting warranty reports and holding parts for warranty purposes
is contrary to the primary mission of field units--to repair the
equipment as soon as possible so that the equipment and the unit can
resume its mission.  A warranty functions contrary to the primary
mission by requiring maintenance personnel to hold parts until a
determination can be made as to whether the part is warranted and how
it should be repaired. 

In addition, maintenance personnel sometimes replace broken parts on
one system with good parts from one or more other systems to keep the
maximum number of weapon systems operating and available, thereby
fulfilling their primary mission.  As a consequence, the broken or
defective parts are moved from their original weapon system.  This
can void a warranty, which can require that the part submitted for a
warranty claim come from the original weapon system that the
contractor delivered to the service. 

The Tank-Automotive and Armaments Command official responsible for
their cost-effectiveness analyses said that historically contractors
only accept about 30 percent of potential claims.  As a result of the
claims submission problem, the Tank-Automotive and Armaments Command
is primarily obtaining systemic warranties instead of threshold
warranties.  However, according to this official, the Tank-Automotive
and Armaments Command has never successfully filed a systemic
warranty claim, and the probability that claims will be filed under a
systemic warranty is zero. 

A report on the Army's warranty program further supports the claims
submission problem.  "Claims submission from the field is low.  Only
a fraction of the work orders are submitted for claims.  For many of
these, the data are inaccurate and incomplete." The report looked at
several commands and weapon systems and calculated that, at the
Tank-Automotive and Armaments Command, only 537 actual claims were
made out of 8,567 potential claims for 21 contracts.\5

The Air Force faces similar low claims submission problems. 
Officials from the Air Force Materiel Command stated that the lack of
reports filed on warranted items from the field is a serious problem. 
They further stated that the most important mission to field
personnel is to repair the items as soon as possible so that the
aircraft can resume its mission.  Many Air Force warranties rely on
the submission of product quality deficiency reports for filing
claims.  An Air Force Inspector General's report\6

estimated that only 15 to 20 percent of failures are actually
reported on these forms because they are complicated and cumbersome
for maintenance personnel to fill out.  Therefore, the Air Force
estimates that 80 to 85 percent of failures go unreported. 

In addition to the lack of incentives for field staff to track and
report warranty claims, there is a lack of credible data systems and
manpower to administer warranty claims.  One Air Force official
responsible for overseeing warranties at a major command said that
because no system is in place to track the warranties or to process
claims efficiently, administering the program is a "nightmare." The
Deputy Assistant Secretary of the Air Force for Contracting indicated
in 1992 that (1) the problems with warranty administration are not
new and (2) the Air Force does not possess and has not been able to
develop data systems designed to track warranted items.  He added
that the lack of necessary manpower resources in the field and in the
program offices for accomplishing warranty administration compounded
this problem. 


--------------------
\5 Evaluation of the U.S.  Army Warranty Program, U.S.  Army Materiel
Systems Analysis Activity (Technical Report No.  547 Aug.  1992). 

\6 Process Effectiveness Review:  Product Quality Deficiency
Reporting, Office of the Inspector General, Headquarters Air Force
Materiel Command (PN 92-06, April 20 - July 29, 1992). 


      BENEFITS CLAIMED FOR
      WARRANTIES MAY BE
      QUESTIONABLE
-------------------------------------------------------- Chapter 2:1.2

Claimed warranty benefits include providing support that could be
viewed as a form of prepaid maintenance and a process for resolving
product performance problems.  Viewed as prepaid maintenance,
warranties pay contractors a sum of money up front based on an
estimate of the number of defects that the government might claim. 
The contractor keeps the difference between actual claims and the
warranty price as his profit.  If the 1989 and 1990 F-15E contracts
were considered a form of prepaid maintenance, then the government
paid the contractor $12 million and made claims totaling $402,000. 
The contractor kept as profit $11.6 million. 

Further, the warranty provides the government a process for dealing
with the contractor and delineates the contractor's responsibilities. 
However, the value of this process seems to vary from system to
system.  While several program officials told us that the contractors
settled claims and fixed problems much more quickly under a warranty,
other officials said that the penalties to the contractor are low
under a warranty and that contractors routinely dispute government
claims.  According to one Air Force official, the Air Force has had
poor results in getting a return on the claims it has filed. 
Warranty officials stated that contractors often stall and argue
about claims because (1) the contractor asserts that it could have
repaired the item more quickly or efficiently, if government
maintenance personnel repair an item and bill the contractor; (2) the
maintenance personnel did not keep the broken part for contractor
inspection; (3) the contractor may believe that the weapon system was
operated outside the performance parameters to which it was designed
or the maintenance personnel damaged the part using improper
procedures, and (4) the contractor may find that the defective part
has been shifted from the original weapon system in which it was
delivered to the government. 


   WARRANTY WAIVERS ARE NOT USED
   AS EXPECTED
---------------------------------------------------------- Chapter 2:2

Congress included a provision in the warranty law that allows the
Secretary of Defense or his designee (no lower than an Assistant
Secretary of Defense or a military department) to waive the
requirement for a weapon system warranty for either national defense
or cost-effectiveness reasons.  However, since 1985 only 21 waiver
requests DOD-wide have reached the assistant secretary level.  Of
those, 15 have been approved.  The conference report accompanying the
bill repealing the 1984 warranty law and enacting the current section
2403 noted clearly that the House and Senate Committees on Armed
Services did not intend DOD obtain to warranties that are not
cost-effective.  The report stated that "a failure to conduct
cost-benefit analyses and to process waivers where cost-effective
guarantees are not obtainable would defeat the legislative intent of
congressional warranty initiatives."

The majority of program officials we interviewed said that they do
not consider waivers a viable option because of (1) the high
placement of the waiver approval authority required by the warranty
law (2) the potential for negative attention being focused on the
program by these high level officials, and (3) the administrative
burden of processing a waiver request.  The result of this reluctance
to seek waivers is that warranties have become essentially mandatory
for all major contracts.  This was noted by the DOD Acquisition Law
Advisory Panel (the Section 800 Panel) in a January 1993 report where
it stated that "the reluctance of DOD to issue warranty waivers
fosters the use of warranties without regard to their
cost-effectiveness."

Service officials at several major commands and at the assistant
secretary level said that requests for waivers bring unwanted and
often negative attention to an acquisition program.  One service
official stated that there is a definite "stigma" attached to waiver
requests and another referred to it as a "nightmare." Further, waiver
requests impose a significant burden on the program office, which has
to generate all the necessary paperwork, including cost-benefit
analyses, and brief them up the chain of command to the assistant
secretary level, with little or no expectation that a waiver will be
approved. 

As an example, the F-16 program office sought a waiver for the
essential performance warranty of the third F-16 multiyear contract
in November 1991.  The entire process--from the completion of the
cost-benefit analysis and decision to seek a waiver to the rejection
of the waiver request--took 11 months.  This was the first attempt by
this program office to obtain a waiver for the weapon system.  The
second attempt, on the 1994 procurement, was rejected after 8 months. 
The Air Force waiver approval process for the F-16 is shown in figure
2.1. 

   Figure 2.1:  Air Force Waiver
   Approval Process for the F-16

   (See figure in printed
   edition.)

Legend:  SPO-System Program Office; ASC-Aeronautical Systems Center;
FM-Financial Management; AFMC-Air Force Materiel Command;
CBA-Cost-Benefit Analysis; SAF-Secretary of the Air Force;
PEO-Program Executive Office; AF-Air Force; AQ-Acquisition;
AQC-Contracting; GCQ-General Counsel for Acquisition; LL-Legislative
Liaison; LGM-Directorate of Maintenance; AQCS-Programs Division,
PEM-Program Element Monitor. 

Source:  GAO, developed from interviews with Air Force officials. 

The cost-benefit analysis for the third multiyear warranty concluded
that both the design and manufacture as well as the materials and
workmanship warranties would not be cost-effective.  According to an
Air Force program official involved in seeking these waivers, the
program office did not request a waiver on these parts of the
warranty because it believed it would be impossible to get approval. 
In addition, the official said that the program office was certain
that it would be directed to renegotiate these warranties with the
contractor.  Rather than seek a waiver for these warranties, the
program office focused on what it considered its strongest case for a
waiver, the essential performance warranty.  The contractor had
produced approximately 1,500 F-16s when the program office began
seeking a waiver and, according to an official in the program office,
the program office knew how the aircraft would perform and also knew
that the essential performance warranty would provide no benefit to
the government.  The program office therefore sought a waiver, which
was denied because the warranty covered only subsystems rather than
the entire weapon system.  The assistant secretary indicated a
warranty covering the entire weapon system was needed before a
decision on whether to grant a waiver could be made. 

The F-16 program office again sought and was denied a waiver for the
essential performance requirement warranty for the fiscal year 1994
procurement.  According to Air Force officials, a determination was
made at the assistant secretary level that, pursuant to the law, a
front-line fighter (a major weapon system) should have a warranty and
the F-16, because it is a system that has been in production for many
years, should have enough data to craft a valid warranty.  Therefore,
the waiver request was rejected.  The warranty obtained contained no
risk to the contractor because the warranty was tied to performance
measurements that the system had already passed, a fact the
contractor and the government already knew.  The warranty thresholds
have a mission reliability of 90 percent and aircraft availability of
85 percent.  The aircraft achieved a mission reliability rate of 97.2
percent and an aircraft availability rate of 91 percent during the
official measurement period.  This warranty was a warranty in name
only. 

It was clear to us from discussions with several officials that they
believe that obtaining a warranty requires a relatively small amount
of time and effort for the program office compared to the amount of
time and effort required to avoid spending that money by obtaining a
waiver.  In addition, according to officials at a major Air Force
command, contractors are aware that waivers are exceedingly difficult
to obtain and may insist on a high warranty price if the program
office seeks an effective warranty.  This in turn may drive the
program office to obtain reduced warranty coverage to reduce the
warranty price.  Program managers do not request a waiver because
they believe it will not be granted and, consequently, unnecessary or
costly warranties are purchased. 

As required by the Federal Acquisition Streamlining Act of 1994
(P.L.103-355), DOD revised DFARS regarding weapon system warranties. 
The new regulations provide guidelines for contracting officers and
program managers to use when developing and negotiating weapon system
warranties.  These regulations state that the use of a weapon system
warranty may not be appropriate in all situations.  Further, a waiver
should be requested if it is determined that obtaining a warranty is
not cost-effective or is inconsistent with the national defense. 
However, program managers still need to obtain a waiver before
deciding not to obtain a warranty.  That process has not been
affected by the revised regulations. 


   COST-BENEFIT ANALYSES ARE NOT
   ADEQUATE
---------------------------------------------------------- Chapter 2:3

The warranty law requires DOD to obtain a warranty unless the
Secretary of Defense determines that warranty would not be
cost-effective or in the interest of the national defense. 
Applicable regulations (DFARS 246.770-7) require that a cost-benefit
analysis be conducted and documented in the contract file to
determine if the warranty is cost-effective.  The Air Force and the
Army had conducted cost-benefit analyses for 21 of the warranties on
the 30 contracts we reviewed.  The cost-benefit analyses performed,
however, were inadequate because (1) the warranties were often not
separately priced, (2) the government's administrative costs were not
fully included, (3) the analyses assumed all potential defects would
be identified and claims submitted, and (4) they did not include a
present value analysis. 

The Navy conducted only one cost-benefit analysis and has not adhered
to DFARS 246.770-7, which states that "in assessing the cost
effectiveness of a proposed warranty, perform an analysis which
considers both the quantitative and qualitative costs and benefits of
the warranty." The Navy's policy is to obtain what it calls "no-cost"
warranties.  However, no warranty is without cost, and not separately
pricing a warranty does not mean the government does not incur a cost
for the warranty, only that the price of the warranty is built into
the cost of the system.\7


--------------------
\7 Warranty Guidebook, Defense Systems Management College (Oct. 
1992). 


      WARRANTIES OFTEN NOT
      SEPARATELY PRICED
-------------------------------------------------------- Chapter 2:3.1

Contracts in all services are often signed without separately pricing
the warranty.  Of the 38 contracts we reviewed, 24 warranties were
not separately priced.  Instead, the warranty price was included in
the price of the product, thereby making it almost impossible to
perform a realistic cost-benefit analysis.  The services have
different policies on pricing a warranty.  Although the Air Force's
policy since 1994 has been to separately price all warranties, the
Army does not require that the warranty be separately priced, and the
Navy maintains that it is not appropriate to negotiate additional
costs for weapon system warranties.  In addition, we were told by a
DOD official that the government often pays twice for the warranty,
once in the actual price of the product and separately in the price
of a warranty. 


      GOVERNMENT ADMINISTRATIVE
      COSTS NOT INCLUDED
-------------------------------------------------------- Chapter 2:3.2

The services performed cost-benefit analyses for 22 out of the 38
contracts we reviewed.  In general, these cost-benefit analyses do
not appear to fully include the administrative costs paid by the
government for warranty development and administration.  Thus, the
cost element of the analysis is kept artificially low.  However, the
Tank-Automotive and Armaments Command uses an estimate, developed in
the early 1980s, that calculates the administrative costs of
processing each warranty claim as $150.  The Army could not provide
us with a copy of the study from which this figure was obtained. 


      LOW CLAIMS SUBMISSION AND
      ACCEPTANCE
-------------------------------------------------------- Chapter 2:3.3

The cost-benefit analyses generally assume that all or a high
percentage of claims will be made and accepted.  We found that the
U.S.  Army Missile Command and the U.  S.  Army Tank-Automotive and
Armaments Command sometimes made greatly different claims submission
and acceptance assumptions when analyzing the costs and benefits of
warranties.  The 1989 Multiple Launch Rocket System cost-benefit
analysis by the Missile Command assumed claims would always be filed
and accepted when warranted items break. 

The Tank-Automotive and Armaments Command official responsible for
its cost-effectiveness analyses used a range of 20 to 90 percent for
the probability that claims would be filed.  As discussed previously,
Army and Air Force studies indicate that actual claims submission
from the field is low.  In the case of the Air Force, possibly as low
as 15 to 20 percent of actual failures. 


      PRESENT VALUE ANALYSIS
      EXCLUDED
-------------------------------------------------------- Chapter 2:3.4

Nine of 12 cost-benefit analyses we sampled did not conduct a present
value analysis as part of the cost-effectiveness review.\8
Cost-benefit analyses normally involve comparing different costs
incurred at different times.  For two or more alternatives to be
compared on an equal economic basis, it is necessary to consider the
costs of each alternative currently or at their "present values."
This recognizes that money has earning power over time.  A present
value analysis is important because it allows the comparison of
current expenses with expected future benefits by taking into account
the time value of money.  Without a present value analysis, comparing
a contract with and without a warranty cannot be done because the
stream of dollars involved are not comparable. 


--------------------
\8 One of the cost-benefit analyses did conduct a present value
analysis and two did not have sufficient documentation to make a
determination. 


   WARRANTY POST-AWARD ASSESSMENTS
   ARE GENERALLY NOT BEING
   PREPARED
---------------------------------------------------------- Chapter 2:4

Our review indicated that the services had not prepared post-award
assessments in 35 of 38 warranties that we reviewed.  There are two
types of post-award assessments required, an in-process assessment
and a final payoff assessment.  The in-process assessments evaluate
whether the claims made under a warranty justify its cost and
document the desirable and undesirable warranty provisions and tasks
for follow-on procurements.  A final payoff assessment evaluates the
economic benefits derived from the warranty compared to the cost of
corrective actions had there been no warranty.  Army regulations
specifically require an in-process and final payoff assessments.  Air
Force regulations only require annual assessments and specify how the
assessments should be performed.  Navy regulations only require that
data be collected to perform an annual assessment of warranty
activity but does not require a final payoff assessment. 

We found that none of the services had prepared the annual in-process
assessments.  In two cases, the Army did prepare final payoff
assessments, it concluded in one that the 1985 Multiple Launch Rocket
System warranty cost the government $1,458,040 more than the benefits
it provided and therefore the warranty was ineffective.  In the other
case, the Army decided that the 1987 Multiple Launch Rocket System
warranty, while appropriate, might have negatively affected the
reliability of the system.  According to the Army Audit Agency,

     Cost-effectiveness analyses and final payoff assessments provide
     the management tools and internal controls that are essential
     and critically needed to make sure the intent of the law is
     satisfied in a way that adequately protects the interests of the
     Army.  Without cost-effectiveness analyses and warranty
     assessments, there is little assurance that the warranties
     obtained and the associated costs were commensurate with the
     benefits received. 

According to Air Force officials with one program office, their
office knows it will have to obtain warranties "no matter what," so
there is no reason for a post-award assessment.  These officials were
referring to the difficulty of receiving a waiver from the
requirement to obtain a warranty, discussed previously. 

For the 18 Army contracts we reviewed, the Army either ignored the
requirement\9 to conduct the final payoff assessments or ignored
findings that showed the warranty benefits did not justify the costs. 
For example, the 1987 Multiple Launch Rocket System final payoff
assessment specifically cited the fact that the thresholds were set
so that the government would repair the first four failures on each
launcher, but only about one claim was actually filed per launcher. 
Because the performance thresholds were set so high, the warranty was
unlikely to serve as an incentive to the contractor to improve the
system's reliability and the warranty may have had a negative effect
on reliability.  The Army obtained a follow-on warranty. 

Air Force regulations covering post-award assessments require the
program manager to monitor warranty feasibility and
cost-effectiveness using annual warranty activity reports submitted
by the contractor or the government.  These assessments are to
include a remarks section that "identifies the warranted tasks or
services that are considered desirable or undesirable based on the
claim frequency, failure mode, and dollar value." The Air Force had
not completed this annual assessment on any of the 12 Air Force
contracts we reviewed. 

In 1987, the Navy\10 issued instructions on warranties and stated
that the Chief of Naval Operations will develop a system for
collecting and analyzing actual warranty use and claim data on an
annual basis.  To date, the Navy has not approved a warranty
information system.  We found only one of the eight naval systems we
reviewed had performed a post-award assessment. 

Air Force program officials stated that final payoff assessments and
post-award assessments are difficult to perform for two reasons. 
First, the warranty is not always separately priced.  Second, the
weapon system warranted may have had many engineering changes from
the time the contract was initially signed until the end of the
warranty period.  These changes make comparing expected costs and
benefits to actual costs and benefits difficult because the initially
projected and actually produced weapon systems are different. 


--------------------
\9 Army Regulation 700-139 (Apr.  10, 1986). 

\10 Secretary of the Navy Instruction 4330.17 (Sept.  1987). 


   CONCLUSIONS
---------------------------------------------------------- Chapter 2:5

Weapon system warranties are generally not cost-effective.  They have
resulted in a significant cost to the government that substantially
exceeds their benefit.  The necessity of negotiating and
administering the warranties also imposes a large, but unquantified
burden on the services. 

The waiver process has resulted in a system in which warranties are
virtually mandatory.  In this system, the program office seeking a
waiver must demonstrate why a warranty would not be cost-effective
and seek approval from an assistant secretary.  It is easier and less
disruptive for that program office to obtain a warranty, regardless
of whether it is necessary or cost-effective, than it is to seek a
waiver.  Because the waiver process is so burdensome and protracted,
warranties are obtained without regard to their cost-effectiveness
and the officials in the program offices have no incentive to conduct
rigorous cost-benefit analyses.  In addition, post-award assessments
have little value to program officials as tools to identify desirable
and undesirable warranties for future contracts because they believe
warranties will have to be obtained "no matter what."

The current DFARS revision, which stresses that weapon system
warranties may not be appropriate in all situations, is a step in the
right direction.  However, it is inadequate to resolve the
difficulties in obtaining a waiver because the regulation could not
change the high level required for approving waivers.  The waiver
approval authority is stipulated in the law itself, and the
incentives that arise from it could not have been changed by this
revision.  As DOD and Congress proceed with acquisition reform, we
believe they need to reexamine the need for and practical
implementation of weapon system warranties. 


WARRANTIES OFTEN NOT APPROPRIATE
FOR USE IN WEAPON ACQUISITIONS
============================================================ Chapter 3

Requiring the routine use of warranties in weapon system acquisitions
is often not appropriate and does not provide the government much in
the way of benefits.  The Institute for Defense Analyses has
identified three functions of a warranty in weapon system
acquisitions--insurance, assurance-validation, and incentivization.\1
In the commercial marketplace, warranties have similar functions. 
Commercial buyers believe warranties protect them against
catastrophic financial losses and excessive operating costs through a
warranty's insurance aspect.  A warranty may also indicate to a buyer
that a product is of better quality, which can be equated to the
assurance-validation function,\2 and may motivate the contractor to
maintain product quality, which equates to the incentivization
function.\3 However, these functions are not as significant in weapon
system acquisitions as they are in buying a commercial product on the
open market. 

For insurance to be cost-effective to the buyer, the risk must be
shared and spread over many insured customers, which is not the case
in weapon acquisitions.  DOD is the only buyer of most weapon systems
and must pay the full cost of the insurance provided by the warranty. 
Also, DOD already has quality assurance processes built into its
contracts to ensure that the product complies with all contract
specifications.  This lessens the need for a warranty's
assurance-validation function.  Finally, in our review of 20 weapon
systems, we could not find any evidence that indicated that a
warranty was a factor in improving system reliability.  Since none of
the traditional benefits conferred by warranties apply to weapon
system purchases, the main benefit that warranties seem to provide is
the extension of the time period DOD has to identify defects to be
corrected by the contractor. 


--------------------
\1 Robert E.  Kuenne, Paul H.  Richanbach, Frederick R.  Ridell, and
Rachel Kaganoff, Warranties in Weapon System Procurement:  Theory and
Practice, Westview Press (1988). 

\2 Lt.  Col.  John P.  Clark, USAF, Report of the Working Group to
Define an Implementable Warranty Program, U.S.  Air Force. 

\3 Wallace R.  Blischke and D.N.  Prabhakar Murthy, Warranty Cost
Analysis, Marcel Dekker, Inc.  (1994). 


   INSURANCE ASPECT IS NOT
   TYPICALLY A GOOD REASON FOR
   OBTAINING A WARRANTY
---------------------------------------------------------- Chapter 3:1

Commercial buyers are interested in stabilizing their operating costs
and protecting themselves against catastrophic losses.  A
manufacturer's warranty provides a commercial buyer with a measure of
insurance against the risks of repair or replacement costs.  If a
warranted product does not perform as specified, the buyer whose
product failed does not face a total financial loss. 

The concept of insurance is based on the principle of shared risk. 
In the commercial marketplace, the cost of offering a warranty is
shared by many buyers who individually pay a small amount of the
total warranty as part of the product's price.  Manufacturers
generally estimate how many of their products will be defective and
price the product to cover this risk.  However, because DOD is
usually the only buyer for a weapon system, the contractor cannot
allocate the cost of insuring against that risk among multiple
buyers.  The complete cost for the estimated risk must be borne by
the sole buyer or absorbed by the contractor.  If it is borne by the
buyer it becomes the price of the warranty.  If it is absorbed by the
contractor then it is a cost that must be covered by the price of the
system.  In both cases the buyer pays.  A further factor in the cost
of a warranty is the extent of unproven technology or innovative
design that a weapon system encompasses, which may cause the
contractor to perceive its financial risk is significant.  This will
tend to drive up the warranty price to the government and causes the
contractors to try to limit warranty coverage as much as possible. 
As a result, insuring against weapon system failures generally is not
beneficial to the government since the government will be responsible
for 100 percent of the estimated cost of that risk.  The government
will only achieve a positive financial result if failures in the
system substantially exceed the contractor's estimates of risk. 

While it may seem that a warranty would make sense in cases where the
extent of system failures exceeds the cost of the warranty, this
occurs in very few instances and the cost of insuring all weapon
systems to cover costs in these instances is not a good financial
decision.  It is for this reason that the government maintains a
policy of self- insurance against losses in almost all other areas. 
A RAND study reported that shifting financial risk to the
manufacturer is seldom an appropriate or sufficient rationale for
obtaining a weapon system warranty, for the same reason that it is
not to the government's advantage to buy insurance from a commercial
firm.\4

Another factor limiting the potential insurance benefits of
warranties is the relationship of the government to major defense
contractors.  If a contractor were someday to incur large losses as a
result of a warranty on a weapon system, the threat of insolvency
might cause the government to excuse the contractor from its warranty
obligation.  Historically, when a defense firm incurs large losses
because of a contract, the government has taken action to provide
relief and prevent the firm from going out of business.  DOD's
ability to provide extraordinary contractual relief to a defense
contractor is recognized in Public Law 85-804.  From 1959 to 1993,
DOD used this provision to provide over $4.3 billion (in constant
1996 dollars) in relief to assist contractors in recovering from
losses.\5 As a result, the utility of warranties may be limited to
collecting small or marginal amounts from a contractor rather than
making good a catastrophic loss. 


--------------------
\4 James P.  Stucker and Giles K.  Smith, Warranties for Weapon: 
Theory and Initial Assessment, RAND Corp.  (N-2479-AF, Apr.  1987). 

\5 David V.  Anthony and Carl L.  Vacketta eds.  Extraordinary
Contractual Relief Reporter.  Federal Publications, Inc.  (1996). 


   WARRANTIES ARE NOT GENERALLY A
   GOOD MEANS OF ASSURING QUALITY
   SYSTEMS
---------------------------------------------------------- Chapter 3:2

The second purpose of warranties is assurance-validation.  Applied to
weapon systems, assurance-validation means assuring DOD that the
manufacturer's product conforms to the design, quality, and
performance levels specified in the contract.  Although DOD is moving
more toward a commercial acquisition system, it uses many quality
assurance processes to verify product quality and contract
conformance independent of the warranty clauses.  Since the cognizant
contract administration office is responsible for verifying that the
accepted product conforms to the specifications in the contract, the
assurance-validation function of a weapon system warranty may not be
necessary. 

In weapon system acquisitions, DOD uses many different program
management tools to reduce the inherent risks of the acquisition
process.  DOD's weapon acquisition policies seek to reduce the risk
of obtaining a poor quality product by establishing a disciplined
multiphased process that (1) translates mission needs into stable and
affordable programs; (2) acquires quality products; and (3) provides
a program management structure that has clear lines of
responsibility, authority, and accountability.  As each weapon system
progresses through the phases, it is subject to comprehensive
programmatic reviews.  During the reviews, an assessment is made of
the program's accomplishments to date, plans for the next phase, and
acquisition strategies for the remainder of the program. 
Additionally, the program risks and risk management planning is
evaluated.  In theory, for each of the acquisition phases,
program-specific results are required before a program is permitted
to proceed to the next phase.  For example, a program may be required
to demonstrate the maturity of a manufacturing process before being
permitted to start production.  Techniques available to manage risk
include the use of technology demonstrations and prototyping to test
hardware, software, manufacturing processes and/or critical
subsystems.  Another technique is to test and evaluate the weapon
system or its components to determine system maturity and identify
technical risks. 

Finally, DOD establishes quality assurance programs to provide
confidence that a weapon system will conform to the technical
requirements and provide satisfactory performance.  A warranty is one
more management tool at DOD's disposal to assure quality as the
weapon system begins to be put in use.  Given the extensive quality
assurance efforts made over the whole development and production
cycle, however, a warranty may actually be insurance against the
failure of the quality assurance process. 


   WARRANTIES MAY NOT PROVIDE MUCH
   INCENTIVE TO IMPROVE PRODUCT
   QUALITY
---------------------------------------------------------- Chapter 3:3

Finally, warranties are supposed to serve as an incentive to
manufacturers to improve quality.  Conceptually, all warranties
motivate a manufacturer to improve product quality because the goal
is to maximize profits by not having to perform warranty service. 
However, when a commercial manufacturer provides a warranty, it
generally knows the projected failure rates of the product and the
probable repair costs.  The costs for these failures are included in
the cost of the product.  Efforts to keep commercial prices
competitive probably undercuts the incentive to make additional
profit by pricing the warranty higher than its expected cost. 

In the commercial market, a warranty may also signal product quality
to the buyer because the commercial buyer generally is not familiar
with how the product was made and does not know how the product will
perform.\6 Therefore, the warranty becomes a marketing tool to help
sell the product by convincing the buyer it is a superior product.\7

In weapon acquisition, the contractor is generally not faced with
direct price competition.  Absent direct price competition, there is
no incentive for the contractor to limit its ability to fully cover
the estimated cost of risks of system failure being warranted. 
Further, the marketing aspects of a warranty are for the most part
irrelevant to DOD.  The warranty's indication of manufacturer
confidence in product quality is not needed because, as shown in the
previous section, DOD is not a typical consumer.  DOD is
knowledgeable about how the weapons it obtains are made and in some
cases helped to design the system.  Generally, DOD knows how the
product will perform and uses other management techniques to maintain
product quality. 

In addition, the majority of officials in the weapon system program
offices we visited stated that either (1) the warranties had not
induced the contractors to take actions to improve the quality of
their warranted products or (2) if there were quality and reliability
improvements due to the warranty, those improvements were marginal
and unmeasurable.  For example, the F-15 warranty manager stated that
the contractor has not designed components of the weapon system
differently nor is the contractor building the weapon system
differently because of the warranty.  The F-110 engine warranty
manager also said that the warranty has not helped to improve the
reliability of the engine. 

In a report issued in August 1992, the U.S.  Army Materiel Systems
Analysis Activity, reached a similar conclusion regarding different
warranties on several systems.  The report stated:  "It is extremely
unlikely that hardware improvements that were performed under the
warranty would not have been performed had there not been a warranty. 
Therefore, it is unlikely that any reliability growth from these
improvements could be attributed to the warranty." The report stated
that no models were available that could be used to measure
reliability improvements resulting from a warranty. 


--------------------
\6 Nancy A.  Lutz, The Economic Theory of Warranties, Virginia
Polytechnic Institute and State University (Jan.  1993). 

\7 A warranty, in the commercial market, also limits the
manufacturers' liability for loss or damage caused by a product's
malfunction or defect.  Usually, the maximum recoverable by a buyer
under a warranty is the full price of the item. 


   WARRANTIES MAY CONFER LIMITED
   BENEFITS AT GREAT EXPENSE
---------------------------------------------------------- Chapter 3:4

With or without a warranty, a contractor is obligated to produce a
product that complies with the terms delineated in the contract,
including all design, manufacturing, and performance specifications. 
During DOD's inspection and quality assurance processes, the
government needs to identify defects or deficiencies and notify the
contractor of problems at the earliest reasonable time.  If DOD did
not obtain a warranty, the contractor could be released from any
further obligation to correct problems once the government accepted
the product.\8 A warranty extends beyond acceptance the period during
which the government can identify defects and require the contractor
to correct them at no charge. 

Prior to acceptance, if defects are discovered, the government, even
without a warranty, can (1) order the contractor to correct the
defects at no additional charge, (2) reject the nonconforming
product, (3) terminate the contract for default, or (4) seek a price
reduction.  A warranty obligates the contractor to correct defects,
even if the government did not identify them before acceptance.  This
can reduce disputes because the warranty eliminates the need to prove
where and when a defect came into existence.  For example, in the
case of a failure-free warranty, the government merely needs to
demonstrate that an item does not work. 

A warranty also supplements inspection by providing the opportunity
to observe the product's performance during a period of use when
additional problems may become apparent.  However, a properly
structured test program could identify such problems early in the
acquisition process, when it is less costly to address
deficiencies.\9

In a contract without a warranty, the government's acceptance of the
product generally ends the contractor's obligation.\10 Even without a
warranty, however, the government can revoke its acceptance and hold
the contractor financially accountable if latent defects are
discovered.  In cases of latent defects, the government must
demonstrate that the defects existed at the time of delivery, but
could not have been discovered by a reasonable inspection. 


--------------------
\8 Acceptance does not relieve the contractor of its obligation to
correct latent defects.  Kaminer Const.  Co v.  United States, 488 F. 
2d 980 (Ct.  Cl.  1973). 

\9 Michael S.  Bridgman and David V.  Glass, Better Assessment of
Operational Suitability, Logistics Management Institute (Jan.  1992). 

\10 The government may seek financial recovery if a contractor
engaged in fraud, or made gross mistakes amounting to fraud.  In
cases of fraud, the contractor willingly and knowingly gives the
government a inferior product.  To recover for fraud, the government
must prove intent to defraud.  Bar-Ray Products, Inc.  v.  United
States, 340 F.  2d 343 (Ct.Cl.  1964).  In the case of gross
mistakes, the government must show deviation from contract
specifications, and misrepresentation of its performance led the
government to accept a non-conforming or defective product. 
Catalytic Engineering and Manufacturing Corp., 72-1 BCA ï¿½ 9342
(1972).  These cases are rare and very difficult to prove. 


   CONCLUSIONS
---------------------------------------------------------- Chapter 3:5

While warranties may have value to a consumer in the commercial
world, obtaining a warranty for a weapon system may be a flawed
concept because (1) the government does not need the insurance
coverage provided by a warranty and cannot share the expense of the
warranty with other customers; (2) warranties are an expensive way to
assure the quality of a weapon system; (3) DOD's quality assurance
activities, should provide much greater assurance of compliance with
contract specifications than does a warranty; and (4) warranties may
not cause contractors to improve the quality of the weapon systems'
they produce.  Further, weapon system warranties provide very limited
benefits to the government.  The only measurable benefit of a
warranty is the ability to have the contractor correct defects for
some negotiated period after acceptance of the product, and as
discussed in the prior chapter this comes at a high cost.  We believe
warranties should be used judiciously and only in cases where their
cost-effectiveness can be clearly demonstrated. 


   RECOMMENDATIONS TO THE
   SECRETARY OF DEFENSE
---------------------------------------------------------- Chapter 3:6

We recommend that the Secretary of Defense establish an expedited
waiver process that limits the disincentives inherent in the current
process.  We also recommend that the Secretary of Defense revise
DOD's acquisition policies to adequately manage those warranties that
the military services determine should be obtained.  Consideration
should be given to (1) requiring that all weapon system warranties be
separately priced in order to allow meaningful cost-benefit analyses;
(2) improving cost-benefit analyses through more realistically
reflecting the likelihood of claim submission, performing present
value analyses, and including the government's administrative costs;
and (3) ensuring that the services enforce the regulations requiring
post-award assessments of weapon system warranties so that the
services will know why these warranties were or were not beneficial
to the government.  We also recommend that the Secretary of Defense
direct the Secretaries of the Air Force and the Navy revise their
regulations to require a final payoff assessment for weapon system
warranties as the basis for purchasing more beneficial follow-on
warranties and building institutional knowledge for procuring and
administering effective warranties. 


   RECOMMENDATION TO CONGRESS
---------------------------------------------------------- Chapter 3:7

The administrative problems that we have identified appear to be
unintended consequences of the warranty law due to the de facto
mandatory nature of warranties.  Attempts to administratively correct
the problem have not been very successful.  Since DOD continues to
have problems administering weapon system warranties and the
warranties provide minimal benefits for the costs incurred, Congress
should repeal 10 U.S.C.  2403.  Were the warranty requirement
repealed, DOD and the services would still have management
flexibility to obtain warranties for major weapon systems only when
deemed appropriate.  As was done prior to the warranty law, DOD and
the services would rely on the FAR and their own policies to
determine when it is appropriate to obtain a weapon system warranty. 
The decision should be documented as part of the system acquisition
strategy. 


   AGENCY COMMENTS AND OUR
   EVALUATION
---------------------------------------------------------- Chapter 3:8

In commenting on a draft of this report, DOD stated that it "strongly
supports" our recommendation that Congress should repeal
10 U.S.C.  2403.  Since 1992, DOD has supported the need for
congressional repeal of the weapon system warranty law.  DOD only
partially concurred with the recommendations to the Secretary of
Defense, stating that the solution to the problems we cited is repeal
of the law.  DOD indicated that it will ask the military departments
to review their warranty waiver process.  DOD noted, however, that in
order to remove the disincentives and streamline the waiver process
it needs relief from the congressional notification requirement and
the warranty waiver approval level. 

DOD stated that it did not see the need to separately price
warranties because it has insight into warranty costs through cost
reporting and can project warranty cost from actual claim data. 
Although DOD currently has insight into warranty costs, we found that
this information is often not used in the cost-benefit analyses. 
Therefore, we believe that separately pricing the warranty would
permit DOD to perform better warranty cost-benefit analyses. 

DOD stated that all the military departments have in place
regulations that require post-award assessments, but acknowledged
that the military departments were not fully complying with existing
regulations.  DOD stated that it would reiterate the importance of
such assessments in a memorandum to the military departments.  Our
review indicated, however, that only the Army's regulation
specifically requires a final pay-off assessment to determine the
economic benefit derived from a warranty.  We believe that the Air
Force and the Navy regulations should be revised to explicitly
require final payoff assessments.  DOD's comments are reprinted in
their entirety in appendix II. 


QUANTIFIED COSTS AND BENEFITS OF
WARRANTIES
=========================================================== Appendix I

                         (Dollars in millions)

                                        Warran  Warran          Warran
                        Contract            ty      ty          ty
Weapon system           number           price  claims     Net  status
----------------------  --------------  ------  ------  ------  ------
F-15E                   F33657-89-C-    $6.000  $0.048  ($5.95  Expire
                        2000                                2)  d

                        F33657-90-C-     6.000   0.354  (5.646  Expire
                        2319                                 )  d

F-16a                   F33657-84-C-    27.860   2.780  (25.08  Expire
                        0247                                0)  d

                        F33657-88-C-    15.100   1.091  (14.00  Majori
                        0037                                9)  ty
                                                                expire
                                                                d

C-17b                   F33657-89-C-    11.563   0.186  (11.37  Expire
                        0001                                7)  d

                        F33657-92-C-    10.500   0.218  (10.28  Expire
                        0030                                2)  d

                        F33657-92-C-    15.471   0.276  (15.19  Expire
                        0031                                5)  d

Multiple Launch Rocket  DAAH01-85-C-     1.584   0.126  (1.458  Expire
System                  1134                                 )  d

======================================================================
Total                                   $94.07  $5.079  ($88.9
                                             8             99)
----------------------------------------------------------------------
\a Final figures for the Multiyear II warranty's claim data were
unavailable because the program office never collected the final
numbers.  Estimates from a 1991 study put the total potential
warranty claims at $9.94 million.  The $2.78 million figure provided
to us is the only figure available.  The essential performance
requirement warranty for the Multiyear III contract has expired and
the materials and workmanship/design and manufacture warranties are
approximately 80 percent expired, according to the program office. 

\b Claim figures for the C-17 are actuals through December 31, 1995. 
They do not reflect the final numbers.  Potential claims may occur as
a result of durability, reliability, maintainability, and
availability testing. 




(See figure in printed edition.)Appendix II
COMMENTS FROM THE DEPARTMENT OF
DEFENSE
=========================================================== Appendix I



(See figure in printed edition.)



(See figure in printed edition.)



(See figure in printed edition.)


The following are GAO's comments on the Department of Defense's (DOD)
letter dated April 22, 1996. 


   GAO COMMENTS
--------------------------------------------------------- Appendix I:1

1.  Separately pricing the warranty would permit improved analysis of
the benefits derived from warranty coverage.  At a minimum, the
negotiated warranty price should be noted in the contract file and
used in the cost-benefit analysis.  Further, the Defense Federal
Acquisition Regulation Supplement (DFARS) requires that the warranty
cost be included as part of the cost-benefit analysis. 

2.  Current Air Force and Navy regulations do not clearly require a
final payoff assessment.  The regulations should be revised to
clearly require a final payoff assessment.  We recognize that the
acquisition community is downsizing and that additional regulations
will not guarantee the assessments will be conducted; however, DOD
and the military departments have a responsibility to properly
execute weapon system contracts.  Part of the responsibility involves
building institutional knowledge on how to procure and administer
weapon system warranties in order to ensure that the government is
obtaining the best value for its procurement dollars. 


MAJOR CONTRIBUTORS TO THIS REPORT
========================================================= Appendix III


   NATIONAL SECURITY AND
   INTERNATIONAL AFFAIRS DIVISION,
   WASHINGTON, D.C. 
------------------------------------------------------- Appendix III:1

David B.  Best
Thomas J.  Denomme
Brad H.  Hathaway
Michele Mackin
Charles W.  Perdue
Jose A.  Ramos


   OFFICE OF THE GENERAL COUNSEL
   WASHINGTON, D.C. 
------------------------------------------------------- Appendix III:2

Margaret Armen
John Brosnan
William Woods


   ATLANTA FIELD OFFICE
------------------------------------------------------- Appendix III:3

Barbara H.  Haynes
John W.  Randall, Jr


   CHICAGO FIELD OFFICE
------------------------------------------------------- Appendix III:4

Michael J.  Jones
Gilbert W.  Kruper


   KANSAS CITY FIELD OFFICE
------------------------------------------------------- Appendix III:5

Charles O.  Burgess
William H.  Gansler


   LOS ANGELES FIELD OFFICE
------------------------------------------------------- Appendix III:6

Dorian R.  Dunbar
Noel J.  Lance

*** End of document. ***