National Missile Defense: Even With Increased Funding Technical and
Schedule Risks Are High (Letter Report, 06/23/98, GAO/NSIAD-98-153).

Pursuant to a congressional request, GAO reviewed the National Missile
Defense (NMD) program funding requirements and schedule and technical
risks, focusing on: (1) why the Department of Defense (DOD)
significantly increased the program's near-term funding in its May 1997
Quadrennial Defense Review; (2) how funding increases authorized and
appropriated by Congress for the program in fiscal years 1996 through
1998 have been used or are planned to be used; and (3) DOD's planned
level of future funding for the NMD program and planned uses for those
funds.

GAO noted that: (1) DOD significantly increased its NMD funding
requirements in May 1997 because more rigorous cost estimates, based on
more detailed program requirements and plans, showed that the program
could not be accomplished within previously projected funding levels;
(2) the 3+3 NMD program was not sufficiently defined for detailed cost
estimating when it initially changed from a technology readiness program
to a deployment readiness program, and was designated a major defense
acquisition program in April 1996; (3) the May 1997 Quadrennial Defense
Review included the first program estimate based on detailed system
descriptions, requirements, and plans; (4) funding increases provided by
Congress in fiscal years 1996 through 1998 were used for risk reduction
activities, such as: (a) retaining competition in the development of the
exoatmospheric kill vehicle, considered one of the most technically
challenging components of the system; (b) increasing the number of
planned tests; and (c) purchasing additional spare hardware; (5)
Congress increased funding for NMD because of concerns about the
adequacy of funding to support the program; (6) the Ballistic Missile
Defense Organization (BMDO) Director acknowledged in an April 1996
testimony that an additional $350 million a year could be used to reduce
program risks; (7) future NMD funding requirements will depend in large
part on the system design and architecture, and when and where it is
deployed; (8) details on the specific system and location are not
expected for some time; (9) program life-cycle costs ranged from $18.4
billion by fiscal year 2003 to $28.3 billion by fiscal year 2006; (10)
since GAO's December 1997 report, DOD has increased funding and revised
NMD program plans to mitigate schedule and technical risks; (11)
however, program officials told GAO that even with the mitigation
actions resulting from the increased funding, schedule and technical
risks associated with a 2003 deployment remain high; (12) according to a
February 1998 report of a panel of former senior military, government,
and industry officials, successful execution of the 3+3 program on the
planned schedule is highly unlikely; and (13) this panel concluded that
the program would benefit from the earliest possible restructuring to
contain the risk.

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

 REPORTNUM:  NSIAD-98-153
     TITLE:  National Missile Defense: Even With Increased Funding 
             Technical and Schedule Risks Are High
      DATE:  06/23/98
   SUBJECT:  Ballistic missiles
             Weapons systems
             Air defense systems
             Defense contracts
             Defense appropriations
             Future budget projections
             Cost analysis
IDENTIFIER:  DOD National Missile Defense Program
             DOD Quadrennial Defense Review
             BMDO 3 Plus 3 Program
             
******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO report.  Delineations within the text indicating chapter **
** titles, headings, and bullets are preserved.  Major          **
** divisions and subdivisions of the text, such as Chapters,    **
** Sections, and Appendixes, are identified by double and       **
** single lines.  The numbers on the right end of these lines   **
** indicate the position of each of the subsections in the      **
** document outline.  These numbers do NOT correspond with the  **
** page numbers of the printed product.                         **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
** A printed copy of this report may be obtained from the GAO   **
** Document Distribution Center.  For further details, please   **
** send an e-mail message to:                                   **
**                                                              **
**                                            **
**                                                              **
** with the message 'info' in the body.                         **
******************************************************************


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


Report to Congressional Requesters

June 1998

NATIONAL MISSILE DEFENSE -
ACQUISITION STRATEGY RISKY DESPITE
INCREASED FUNDING

GAO/NSIAD-98-153

National Missile Defense Funding and Risks

(707300)


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

  BMDO -
  DOD -
  NMD -

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


B-279053

June 23, 1998

The Honorable Carl Levin
Ranking Minority Member
Committee on Armed Services
United States Senate

The Honorable Jeff Bingaman
Ranking Minority Member
Subcommittee on Strategic Forces
Committee on Armed Services
United States Senate

This report responds to your request that we review the National
Missile Defense (NMD) program funding requirements and schedule and
technical risks.  Specifically, you asked us to determine (1) why the
Department of Defense (DOD) significantly increased the program's
near-term funding in its May 1997 Quadrennial Defense Review,\1 (2)
how funding increases authorized and appropriated by Congress for the
program in fiscal years 1996 through 1998 have been used or are
planned to be used, and (3) DOD's planned level of future funding for
the NMD program and planned uses for those funds.  You also asked for
an assessment of the program's current schedule and technical risks. 
We provided an initial assessment of the schedule and technical risks
in our December 12, 1997, report\2 to you.  This report updates that
assessment. 

You also asked us to review a report covering some of these same
issues that DOD was to provide to the Committee on Armed Services by
February 15, 1998.  We were unable to review that report because DOD
had not released it, as of May 21, 1998. 


--------------------
\1 The Quadrennial Defense Review was commissioned to provide a
comprehensive examination of the defense strategy, force structure,
force modernization plans, infrastructure, budget plan, and other
elements of the defense program and policies. 

\2 National Missile Defense:  Schedule and Technical Risks Represent
Significant Development Challenges (GAO/NSIAD-98-28, Dec.  12, 1997). 


   BACKGROUND
------------------------------------------------------------ Letter :1

The primary mission of NMD is to defend the United States against an
intercontinental ballistic missile attack consisting of several
missiles launched from a rogue nation.  It would also have some
capability against an accidental launch from nuclear powers such as
Russia or China.  The United States has been developing technologies
for use in an NMD system for a number of years.  In April 1996 DOD
changed the purpose of the NMD program from a technology readiness
program to a deployment readiness program and designated NMD as a
major defense acquisition program.  Under the technology readiness
program, the Ballistic Missile Defense Organization (BMDO) developed
and matured technologies for possible use in an NMD system.  Under
the current deployment readiness program, BMDO plans to integrate the
technologies into a system that can be made operational.  The
deployment readiness program is commonly known as the "3+3" program. 

The goal of the NMD 3+3 program is to develop and demonstrate, by
fiscal year 2000, an initial, limited capability that could be
deployed by fiscal year 2003.  The deployment decision is to be based
on ongoing assessments of the threat and will not be made until
fiscal year 2000 at the earliest.  If DOD concludes at that time that
the threat does not warrant deployment by fiscal year 2003,
development will continue. 

While BMDO is still determining the specific design of the initial
NMD system, its features will include (1) space- and ground-based
sensors to provide early warning of attacking missiles; (2)
ground-based radars to identify and track the threatening warheads;
(3) ground-based interceptors to collide with and destroy incoming
warheads; and (4) a battle management, command, control, and
communications system (see fig.  1). 

   Figure 1:  Representative NMD
   System

   (See figure in printed
   edition.)

Source:  DOD

The NMD system architecture will evolve over time through
incorporation of advanced technologies to defend against more
sophisticated threats.  For example, the Space Based Infrared
System--Low Earth Orbit--a group of satellites that will track
incoming warheads and help discriminate between the warheads and
other objects such as decoys or debris--will be added to the system
at a later time.  BMDO also has not determined where any NMD system
would be located.  NMD development is to be conducted within the
terms of the Anti-Ballistic Missile Treaty,\3 but a deployed system
either could be compliant with the treaty as written or might require
amendment of the treaty's provisions, according to DOD officials. 

When the 3+3 program was established in April 1996, DOD estimated
that research and development costs for the period fiscal year 1998
through fiscal year 2003 would total $2.3 billion.  In May 1997 DOD
released the results of its Quadrennial Defense Review, which
estimated research and development costs for fiscal years 1998
through 2003 would total about $4.6 billion--almost twice as much as
the April 1996 estimate. 


--------------------
\3 The Anti-Ballistic Missile Treaty between the United States and
the former Soviet Union governs the conditions under which
anti-ballistic missile systems and components can be developed and
deployed.  The treaty as currently formulated would limit deployment
to a single site near Grand Forks, North Dakota. 


   RESULTS IN BRIEF
------------------------------------------------------------ Letter :2

DOD significantly increased its NMD funding requirements in May 1997
because more rigorous cost estimates, based on more detailed program
requirements and plans, showed that the program could not be
accomplished within previously projected funding levels.  The 3+3 NMD
program was not sufficiently defined for detailed cost estimating
when it initially changed from a technology readiness program to a
deployment readiness program and was designated a major defense
acquisition program in April 1996.  The May 1997 Quadrennial Defense
Review included the first program estimate based on detailed system
descriptions, requirements, and plans. 

Funding increases provided by Congress in fiscal years 1996 through
1998 were used for risk reduction activities, such as retaining
competition in development of the exoatmospheric kill vehicle,\4
considered one of the most technically challenging components of the
system; increasing the number of planned tests; and purchasing
additional spare hardware.  Congress increased funding for NMD
because of concerns about the adequacy of funding to support the
program.  The BMDO Director acknowledged in April 1996 testimony that
an additional $350 million a year could be used to reduce program
risks. 

Future NMD funding requirements will depend in large part on the
system design and architecture and when and where it is deployed. 
Details on the specific system and location are not expected for some
time.  To identify possible future funding needs, BMDO has estimated
four different scenarios.  Program life-cycle costs\5 associated with
these scenarios ranged from $18.4 billion for a deployment at Grand
Forks, North Dakota, by fiscal year 2003 to $28.3 billion for a
deployment by fiscal year 2006. 

Since our December 1997 report, DOD has increased funding and revised
NMD program plans to mitigate schedule and technical risks.  However,
program officials told us that even with the mitigation actions
resulting from the increased funding, schedule and technical risks
associated with a 2003 deployment remain high.  According to a
February 1998 report of a panel of former senior military,
government, and industry officials, successful execution of the 3+3
program on the planned schedule is highly unlikely.  This panel
concluded that the program would benefit from the earliest possible
restructuring to contain the risk. 


--------------------
\4 The exoatmospheric kill vehicle is the front end of the
ground-based interceptor that will see the target and destroy it by
colliding with it, outside the atmosphere. 

\5 Life-cycle costs include costs to develop and produce system
components, construct facilities, deploy the system, and operate it
for 20 years. 


   NMD PROGRAM NOT SUFFICIENTLY
   DEFINED FOR RELIABLE COST
   ESTIMATE AT THE TIME 3+3
   PROGRAM ESTABLISHED
------------------------------------------------------------ Letter :3

The NMD cost estimate has evolved as the system requirements and
program plans have become better defined.  When the NMD program was
changed to a deployment readiness program in April 1996, plans and
requirements were not sufficiently defined to allow the development
of a reliable cost estimate.  Fiscal year 1996 and 1997 budget
requests were submitted to Congress before the program was changed to
a deployment readiness program.\6

In late 1995 and early 1996, DOD conducted a "Program Update Review"
to determine how to proceed with the NMD program.  The review
considered a number of options for NMD.  The option selected included
an integrated test in fiscal year 1999 and a possible deployment
decision in fiscal year 2000.  DOD estimated that research,
development, and test and evaluation costs for this option would
total about $2.3 billion for fiscal years 1998 through 2003. 
According to program office officials, the update review was based on
a "rough order of magnitude" cost estimate derived from engineering
judgment and field estimates.  Detailed system requirements had not
been established from which to make a formal, documented cost
estimate. 

Once NMD became a deployment readiness program in 1996, the focus
changed from technology and component development to development and
testing of a system that could be quickly deployed.  One of the first
steps was to define operational requirements for the system.  U.S. 
Space Command defined broad requirements for an NMD system in August
1996.  This was followed by NMD's first system requirements review
held in November 1996.  Once these requirements were known, they had
to be defined in sufficient detail so that the contribution of each
system component to the requirement could be determined.  According
to DOD officials, it was only after these detailed requirements were
established that detailed cost estimates could be produced. 

The NMD program office used the requirements data to prepare a new,
more rigorous cost estimate.  DOD's Office of Program Analysis and
Evaluation also prepared an independent cost assessment.  These
estimates were not completed in time to affect the fiscal year 1998
President's budget request.  The program office estimated that about
$4.6 billion would be required for research, development, test and
evaluation--about $2.3 billion higher than previous projections.  The
independent assessment confirmed the program office's projection of
research, development, test and evaluation costs.  As a result of
these estimates, it was apparent to DOD officials that the NMD
program was significantly underfunded.  According to DOD officials,
these were the first disciplined, system-level cost estimates based
on requirements necessary to field an NMD system. 

The Quadrennial Defense Review, which was underway at the time the
estimates were prepared, examined three options for the NMD program: 

  -- The first option was to keep NMD within its current budget,
     which would mean that system deployment would be delayed by at
     least 3 years or that the program would once again become a
     technology readiness program. 

  -- The second option was to increase program funding to the levels
     indicated by the new estimates--an increase of about $2 billion
     in fiscal years 1998 through 2003--in order to maintain the 3+3
     program schedule.  Even with the additional funding, however,
     schedule risks were predicted to remain high. 

  -- The third option was to increase program funding by up to $1.5
     billion but also extending the schedule by about 3 years. 

The review recommended the second option--increased funding to
maintain the option to make a deployment decision in 2000.  The
Secretary of Defense asked Congress to increase the fiscal year 1998
budget request for NMD by $474 million.  Congress appropriated the
requested additional funds.  DOD estimated that an additional $1.8
billion would be needed for fiscal years 1999 through 2003, bringing
the total increase to about $2.3 billion.  The amount of increased
funding was based on the Office of Program Analysis and Evaluation's
independent cost assessment. 

Figure 2 shows the chronology of events leading to the cost estimate
used in the Quadrennial Defense Review. 

   Figure 2:  Chronology Leading
   Up to Cost Estimate Used in
   Quadrennial Defense Review

   (See figure in printed
   edition.)


--------------------
\6 The fiscal year 1996 budget was submitted in February 1995 and the
fiscal year 1997 budget request was submitted in March 1996. 


   CONGRESSIONAL FUNDING INCREASES
   USED FOR RISK REDUCTION
   ACTIVITIES
------------------------------------------------------------ Letter :4

Congress authorized and appropriated significantly more funds in
fiscal years 1996 through 1998 than were requested in the President's
budgets for those years because of its concerns that the budget
requests were not adequate.  DOD officials acknowledged in testimony
that additional funds were needed to reduce program risks.  Funding
increases have been used primarily for risk reduction activities,
such as extending the exoatmospheric kill vehicle design competition
until after competing designs are more fully tested. 


      CONGRESSIONAL FUNDING
      INCREASES
---------------------------------------------------------- Letter :4.1

Because of concerns about the adequacy of funding to support the
deployment readiness program and based on testimony by senior defense
officials, Congress appropriated significantly more funds for the NMD
program in fiscal years 1996 through 1998 than were requested in the
President's budgets for those years.  Figure 3 shows the President's
initial budget requests for research, development, test, and
evaluation funds for fiscal years 1996 through 1998 and the amounts
actually appropriated. 

   Figure 3:  Funding Requested
   and Appropriated for NMD
   Program in Fiscal Years 1996
   through 1998 (Research,
   Development, Test, and
   Evaluation funds only)

   (See figure in printed
   edition.)

Even though no detailed estimate of NMD costs had yet been performed,
senior DOD officials testified that funding in excess of the
President's budget requests would be needed to plan a program with
acceptable risks.  For example, in April 1996, the BMDO Director said
that an additional $350 million above the President's budget requests
in fiscal years 1997 through 1999 could be used to achieve a lower
risk 3+3 program.  He proposed using the additional funding to
increase testing, accelerate development of a booster for the
ground-based interceptor, buy spare hardware to eliminate single
point failures, continue competition between the two kill vehicle
contractors until after at least one additional flight test per
contractor, and increase deployment planning and preparations. 


      DOD USE OF CONGRESSIONAL
      FUNDING INCREASES
---------------------------------------------------------- Letter :4.2

In fiscal years 1996 through 1998, Congress authorized and
appropriated a total of $1,174 million more than the President's
budget requests for those years.  Over 80 percent of additional
funding has been allocated to six program areas--the ground-based
interceptor; ground-based radar; system integration; battle
management, command, control, and communications system; systems
engineering; and test and evaluation.  According to NMD officials,
these funding increases have been or are being used for risk
reduction activities and to execute the 3+3 program.  Table 1 shows
how the funding increases for the 3-year period, fiscal year 1996 to
1998, have been allocated.  Appendix I shows, in detail, the amounts
budgeted and actual or planned program allocations for projects in
fiscal years 1996 through 1998. 



                                Table 1
                
                  Allocations of Congressional Funding
                Increases for Fiscal Years 1996 Through
                                  1998

                                    Funding increase
                                         (dollars in
Program area                               millions)  Percent of total
----------------------------------  ----------------  ----------------
Ground-based interceptor                        $434                37
Systems integration                              159                14
System test and evaluation                       149                13
Ground-based radar                               107                 9
Battle management, command,                       72                 6
 control, and communications
Systems engineering                               57                 5
Other                                            196                17
======================================================================
Total                                         $1,174               100
----------------------------------------------------------------------
Note:  totals may not add due to rounding. 

The largest increase--$434 million or over one-third of the 3-year
total--has been allocated to the ground-based interceptor.  Most of
these funds have been used to maintain competition in the design and
development of the interceptor's kill vehicle.  Original plans were
to select a single kill vehicle design and contractor at the end of
1995 before either of the two competing designs had been fully
tested--even though the kill vehicle is considered one of the most
complex parts of the NMD system.  The additional funding has allowed
the program to preserve the kill vehicle competition through actual
intercept tests in fiscal years 1998 and 1999.  Some of the increased
funding was also needed to cover the costs of a schedule slippage due
to the failure of a flight test in January 1997, purchase a spare
kill vehicle from one of the contractors, and upgrade launch
capabilities at the test range.  Because of subsequent funding
reductions\7

and a decision to incorporate the ground-based interceptor into a
lead system integration contract, program officials decided not to
begin development of a booster for the interceptor. 

Increases totaling $159 million allocated to systems integration have
been used to obtain a prime contractor for the system.  According to
program officials, BMDO decided in the summer of 1996 that a prime
contractor would be needed to manage the remaining design and
development effort and to integrate and test the complete NMD system. 
Two competitive concept development phase contracts were awarded in
fiscal year 1997.  One of two concept development phase contractors,
Boeing North American Company, was selected as the prime contractor
on April 30, 1998. 

An increase of $149 million in the system test and evaluation effort
has been used in part for additional test targets.  Some of the
increased funding has also been used to develop an integrated system
test capability needed for ground tests of the various elements of
the NMD system.  The added funding also permitted increased testing
such as using targets of opportunity to test ground-based system
elements and a Midcourse Space Experiment designed to obtain
information on viewing targets against earth and space backgrounds--a
critical capability in identifying and tracking threatening warheads. 

Funding increases amounting to $107 million allocated to the
ground-based radar have been used to enhance realism in and to
accelerate development of the radar that will be used in testing. 
Original plans were to conduct the tests with a radar technology
demonstrator.  However, with the increased funding, BMDO decided to
construct a ground-based radar prototype to be used in the testing
program.  The prototype has a larger face than the demonstrator and
more closely resembles the radar to be deployed.  Additionally, the
radar development was accelerated. 

After NMD became a deployment readiness program, officials said that
it became apparent that a more extensive battle management, command,
control, and communications effort was needed to support an NMD
system.  This effort is supposed to provide engagement planning and
execution, allow human-in-control of the NMD system, and interface
with external command, control, and communications systems.  With $72
million in additional funding allocated to this element, officials
have been able to begin development of five capability increments of
a prototype battle management, command, control, and communications
system.  The first two increments have been completed and the third
is expected to be completed in April 1998.  Also added was the NMD
communication network and a system that will be used to communicate
with the NMD interceptor in-flight. 

Originally planned funding levels for systems engineering were
sufficient only to support a technology readiness program, according
to program officials.  Funding for this effort was increased by $57
million mostly in order to prepare and update documents required for
a system deployment.  The officials said that without the additional
funding, they would not have been able to baseline the NMD system
architecture, and, thus, there would not be an NMD system. 

The remaining $196 million of the increases was allocated in smaller
increments to a number of areas.  The largest of these was an
increase of about $50 million for program management support.  The
increase paid for personnel and contractor support for the joint
program office as well as for systems analyses and small business
innovative research.  Personnel costs previously spread through all
the projects were rolled up into one project management line item. 


--------------------
\7 Because of changes in inflation assumptions and other cuts
mandated by DOD and Congress in fiscal years 1996 through 1998, NMD
funding was reduced by a total of about $80 million below the amounts
appropriated for the program in those years. 


   FUTURE FUNDING REQUIREMENTS
   DEPEND ON VARIABLES YET TO BE
   CHOSEN
------------------------------------------------------------ Letter :5

Future NMD funding requirements depend in large part on how the
system is designed and when and where it will be deployed.  These
factors may not be known for some time.  For example, the government
and prime contractor have not yet agreed on a final system design. 
The deployment schedule and location will not be known until at least
the fiscal year 2000 deployment review.  To provide a basis for
estimating near-term funding requirements and to help determine how
these differences will impact future funding needs, the program
office prepared four different life-cycle cost estimates, based on
two locations--one at Grand Forks, North Dakota, and the other in
Alaska--and two capability levels--one available in fiscal year 2003
and the other in fiscal year 2006.\8 The life-cycle cost estimates
show the total costs to develop and produce system components,
construct facilities, deploy the system, and operate it for 20 years. 

Figure 4 shows the life-cycle costs estimated for each deployment
alternative.\9

   Figure 4:  NMD Program
   Life-Cycle Cost Estimates

   (See figure in printed
   edition.)

None of the estimates include costs incurred prior to fiscal year
1998.\10

Because specific designs have not yet been determined for system
components, the estimates are based on assumptions about which
designs will be chosen.  The cost estimates could change based on
decisions made by the prime contractor, or evolution of the threat. 

The higher cost for a deployment in Alaska by 2003 is due, in large
part, to the fact that less infrastructure currently exists there,
transportation costs are higher, the construction season is shorter,
and the environment is harsher.  Procurement and operation and
support costs are primarily dependent on the type and amount of
hardware included in the deployment.  Research and development costs
would be slightly higher for an Alaska deployment primarily because
of the need for additional site survey studies. 

The 3+3 program is designed to enable a system to be deployed as
early as fiscal year 2003, but a more capable system could be
operational in fiscal year 2006, according to BMDO.  The primary
differences between the two capability levels used in the cost
estimates are in the type and amount of hardware included.  For
example, the more capable system would have significantly more
interceptors, fewer ground-based radars, but would also include a
space-based sensor system.  After the space-based sensor system is
deployed, fewer ground-based radars will be needed for an Alaskan
deployment because of Alaska's location relative to potential
threats.  The requirement for fewer radars is the primary reason an
Alaskan deployment by fiscal year 2006 is estimated to have a
life-cycle cost slightly less than a deployment at Grand Forks in
that same timeframe.  With fewer radars, operating costs would also
be lower in Alaska. 

The Office of Program Analysis and Evaluation also prepared
independent estimates of NMD program costs in January 1998. 
According to officials responsible for the estimate, costs in the
independent estimates were about 10 percent higher than the estimates
prepared by the program office, due primarily to the fact that the
independent estimates included "pre-planned product improvements" not
included in the program office estimates.  DOD did not provide us
access to the January 1998 independent estimate.  Therefore, we could
not confirm what officials told us about them. 

DOD included funding in its future years defense plan to complete the
research and development phase of the initial capability NMD program
based on the cost estimate used in the Quadrennial Defense Review. 
No funding has been identified for production and deployment of the
system.  According to BMDO, funds for system production and
deployment will not be included until after the fiscal year 2000
deployment decision review.  Figure 5 compares the President's Budget
submissions for fiscal years 1998 and 1999. 

   Figure 5:  NMD Funding
   Estimates Associated with
   President's Budgets for Fiscal
   Years 1998 and 1999

   (See figure in printed
   edition.)

Note:Amounts shown for fiscal years 1996 and 1997 are amounts
actually appropriated. 

Even the fiscal year 1999 funding estimate does not include amounts
that will be needed beginning in fiscal year 2001 to develop system
improvements to keep up with changes in the threat.  According to one
estimate, about $765 million above the President's fiscal year 1999
budget estimate will be needed in fiscal years 2001 through 2003 to
develop the upgrades. 


--------------------
\8 An initial operating capability would be established in fiscal
year 2006, but the full operating capability would not be achieved
until fiscal year 2009. 

\9 Estimates include an allowance for inflation and are intended to
reflect the purchasing power in the year that funds are expended. 

\10 Congress appropriated about $1.6 billion in fiscal years 1996 and
1997 for the NMD program. 


   SCHEDULE AND TECHNICAL RISK
   REMAIN HIGH
------------------------------------------------------------ Letter :6

In December 1997, we reported that DOD faces significant challenges
in the NMD program because of high schedule and technical risk.  We
pointed out that schedule risk was high because the schedule requires
a large number of activities to be completed in a relatively short
amount of time.  Some development activities are not able to proceed
in earnest until the government and prime contractor agree on a final
system design.  Furthermore, developing and deploying an NMD system
in the 6 years allotted under the 3+3 program will be a significant
challenge for DOD given its past history with other weapon systems. 
For example, NMD's acquisition schedule is about one-half as long as
that of the Safeguard system, the only U.S.-based ballistic missile
defense system developed so far.  The program's technical risk is
high because the compressed development schedule only allows limited
testing.  The NMD acquisition strategy called for conducting (1) one
system test prior to the initial deployment decision--a test that
would not include all system elements or involve stressing conditions
such as multiple targets--and (2) one test of the integrated
ground-based interceptor before production of the interceptor's
booster element must begin.  If subsequent tests reveal problems,
costly redesign or modification of already produced hardware may be
required. 

Since our December report, DOD has revised program plans to mitigate
schedule and technical risk to some extent.  Changes include
procuring additional spare hardware to protect against further
schedule slips and increasing the amount of planned testing.  DOD
officials told us, however, that overall schedule and technical risk
associated with a 2003 deployment will remain high, despite these
actions. 


      SCHEDULE RISK
---------------------------------------------------------- Letter :6.1

Even with the additional funding, the program's schedule risk will
remain high, according to DOD officials.  Accomplishing all of the
required contracting, development, integration, and testing planned
before the initial decision point in fiscal year 2000 is, and will
continue to be, high risk.  According to the program manager,
additional funding cannot be used to reduce schedule risk because "we
simply cannot buy back time." However, the additional funds can help
mitigate further slips in the program schedule, according to the
program manager.  For example, additional funds have been identified
to purchase back-up hardware to prevent unnecessarily long delays in
test programs if something goes wrong, as it did in January 1997 when
a test had to be aborted after the target was launched.  That test
could not be repeated for about 6 months due to the lack of a back-up
target. 

In February 1998, a panel of former senior military, civilian, and
industry leaders confirmed our assessment that the 3+3 program
contained high schedule risk.\11 According to the study panel, which
was established by BMDO; the Director of Operational Test and
Evaluation; and the Director of Test, Systems Engineering and
Evaluation; schedule pressures on NMD have created a planning
environment at least as optimistic as that which led to test failures
and delays in other missile defense programs.  In the judgment of the
study panel, successful execution of the 3+3 program on the planned
schedule is highly unlikely.  The panel recommended restructuring the
program to contain the risk and eliminate unrealistic expectations. 


--------------------
\11 Institute for Defense Analyses, Report of the Panel on Reducing
Risk in Ballistic Missile Defense Flight Test Programs, February 27,
1998. 


      TECHNICAL RISK
---------------------------------------------------------- Letter :6.2

Technical risks remain high for a fiscal year 2003 deployment even
though the program has made some technical progress and has revised
plans to increase the amount of testing prior to deployment.  The
amount of flight testing is still limited compared to other programs. 
Other outside reviewers have also commented on the limited amount of
flight testing planned for the program. 

Since our December 1997 report, the program has made some technical
progress.  In January 1998, BMDO conducted its second kill vehicle
sensor test.  An earlier test in June 1997 included a sensor built by
a competing company.  The purpose of both tests was to analyze the
ability of the respective sensors to identify and track objects in
space.  According to DOD, both sensor tests were successful.  The
sensors successfully tracked and obtained data needed to identify
simulated threat targets and decoys.  The two competing contractors
are scheduled to test the ability of their kill vehicle designs to
actually intercept targets in space during fiscal year 1999.  This
data will be used to select a single kill vehicle design and
contractor. 

As a result of added funding, BMDO has also increased the number of
tests planned.  For example, BMDO almost doubled the number of
planned integrated ground tests,\12 added one integrated flight test
prior to the fiscal year 2000 deployment readiness review, and
increased the number of flight tests planned between the readiness
review and the system's initial operational capability date.  The
number of flight tests to be conducted after the readiness review
depends on whether or not a decision is made to deploy.  Without a
deployment decision, there will be two integrated flight tests per
year.  If a deployment decision is made in fiscal year 2000, with a
target deployment of fiscal year 2003, there would be three flight
tests in fiscal year 2000, and four a year in fiscal years 2001
through 2003. 

Overall technical risk associated with a fiscal year 2003 deployment
remains high because the amount of testing, although increased, is
still limited compared to other programs.  Even after the increase in
the number of tests, the program manager told us that in his view,
the planned flight test program is anemic.  The program plans a
maximum of 16-system level flight tests through the end of fiscal
year 2003, the earliest planned deployment date.  By contrast, the
Safeguard\13 program included 111 flight tests before the system
became operational.  Of these 111 tests, 70 were intercept tests, 58
of which were successful.  The panel on reducing risk in ballistic
missile defense programs also concluded that plans for the 3+3
program are based on inadequate test assets and testing.  The panel
recommended increasing the number of tests (both ground and flight
tests) and that the flight test program be restructured to allow more
time between tests to ensure that problems are corrected and the
corrections are tested. 

Technical risk in the NMD program is also of concern to DOD's testing
organization.  According to the Director of Operational Test and
Evaluation's Annual Report for fiscal year 1997, the planned NMD test
program will provide only a limited basis for evaluating system
performance.  The limitations cited in the report include (1) the
limited amount of testing planned prior to the deployment readiness
review; (2) the fact that the booster to be used in the ground-based
interceptor will not be tested prior to the readiness review; (3) the
interface between the system's battle management, command,
communications, and control element and the national command
authority will not be tested before the decision review; (4) the
system's performance against multiple targets will not be tested; and
(5) models and simulations used to support the review will have
minimal validation by real flight data. 

NMD program officials told us that they are in the process of
redefining the program's risk.  The new risk assessment is scheduled
to be completed and documented by early June 1998.  They also pointed
out that the prime contractor's system design and program plans may
impact risk.  According to the program's test and evaluation master
plan, the amount of testing is unlikely to change as a result of
prime contractor selection. 


--------------------
\12 Integrated ground tests will be conducted with certain hardware
and software components integrated into simulations.  Although they
are used to evaluate system performance, they do not include actual
targets or interceptor firings. 

\13 Safeguard, the only other U.S.-based missile defense system,
became operational in 1975.  The program was terminated in 1976. 


   CONCLUSIONS
------------------------------------------------------------ Letter :7

The May 1997 Quadrennial Defense Review recommended significant
increases in NMD funding.  The increases resulted in large part from
a better definition of system requirements and program plans.  As
requirements and plans evolved, estimated development costs almost
doubled.  Increased funds provided by Congress in fiscal years 1996
through 1998 have enabled BMDO to conduct risk reduction activities,
such as purchasing back-up hardware and preserving competition for
the development of the exoatmospheric kill vehicle.  However, despite
the additional activities, the risk of the program being completed on
its current schedule is still high.  Also, any decision in fiscal
year 2000 to deploy an NMD system by 2003 would involve high
technical risk because the associated compressed schedule will permit
only limited testing of the system. 


   AGENCY COMMENTS
------------------------------------------------------------ Letter :8

In commenting on a draft of this report, DOD concurred that the NMD
program "faces significant challenges because of high schedule and
technical risks." DOD also provided technical comments, which we
incorporated as appropriate.  DOD's comments are reprinted in
appendix II. 


   SCOPE AND METHODOLOGY
------------------------------------------------------------ Letter :9

To identify changes that led the May 1997 Quadrennial Defense Review
to recommend significant increases in near-term funding for the NMD
program, we compared documentation available to support the review
estimate to documentation available to support earlier estimates such
as the 1996 Program Update Review.  We discussed differences and
reasons for them with NMD, BMDO, and DOD officials responsible for
estimating NMD program costs. 

To determine how funding increases authorized and appropriated by
Congress for fiscal years 1996 through 1998 were or are planned to be
used, we compared original budgets documents to current funding
allocation documents and discussed reasons for differences with
officials responsible for managing various elements of the NMD
program, such as the ground-based interceptor and test and evaluation
managers. 

To determine the level of future funding and how those funds are to
be spent, we reviewed NMD's cost estimates and assessments of those
estimates.  We discussed the estimates and assessments with NMD and
BMDO officials responsible for preparing them and with DOD officials
responsible for preparing an independent cost estimate.  We did not
review the independent cost estimate because DOD officials did not
provide us with access to it. 

To assess the program's schedule and technical risk, we analyzed the
program's status, strategy for accomplishing the remaining
development work and meeting fielding requirements, and approaches to
demonstrating the system's capabilities and military suitability.  We
also reviewed independent studies of the system's risk and discussed
risk levels and approaches to mitigating risk with NMD program
officials and the program's systems engineering contractor.  We
prepared an initial risk assessment in December 1997 and updated that
assessment for this report. 

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

As agreed with your office, unless you publicly announce its contents
earlier, we plan no further distribution of this report until 15 days
from its issue date.  At that time, we will send copies of this
report to other interested congressional committees, the Secretary of
Defense, and the Directors of the Ballistic Missile Defense
Organization and the Office of Management and Budget.  Copies will
also be made available to others on request. 

If you our your staff have questions concerning this report, please
contact me at (202) 512-4841.  The major contributors to this report
were Lee Edwards, David Hand, Bobby Hall, and Judy Lasley. 

Allen Li
Associate Director
Defense Acquisitions Issues


RESEARCH, DEVELOPMENT, TEST, AND
EVALUATION FUNDS BUDGETED AND
ALLOCATED FOR NMD IN FISCAL YEARS
1996 THROUGH 1998
=========================================================== Appendix I

                              (Dollars in millions)

       Fiscal year 1996       Fiscal year 1997       Fiscal year 1998
     ---------------------  ---------------------  ---------------------
Pro
gra
m
Ele                                                                        Total
men  Budge  Allocat  Chang  Budge  Allocat  Chang  Budge  Allocat  Chang  increa
t        t      ion      e      t      ion      e      t      ion      e      se
---  -----  -------  -----  -----  -------  -----  -----  -------  -----  ------
Sys      0        0      0      0   $24.10  $24.1  $7.09  $141.73  $134.  $158.7
 te                                             0                     64       4
 ms
 in
 te
 gr
 at
 io
 n
Gro  $115.  $255.30  $140.  $147.   272.00  124.1  127.5   297.46  169.9  434.14
 un     22              08     85               5      5               1
 d-
 ba
 se
 d
 in
 te
 rc
 ep
 to
 r
Bat
 tle 33.54    70.02  36.48  32.76    50.65  17.89  43.73    61.67  17.94   72.31
 ma
 na
 ge
 me
 nt
 /
 co
 mm
 an
 d,
 co
 nt
 ro
 l,
 an
 d
 co
 mm
 un
 ic
 at
 io
 ns
Gro  37.78    83.50  45.72  41.50    66.13  24.63  19.54    55.73  36.19  106.54
 un
 d-
 ba
 se
 d
 ra
 da
 r
Upg      0     8.49   8.49   9.35    12.12   2.77  16.75    15.41      -    9.92
 ra                                                                 1.34
 de
 d
 ea
 rl
 y
 wa
 rn
 in
 g
 ra
 da
 rs
Sys  20.80    60.16  39.36  30.83    47.12  16.29  41.94    42.97   1.03   56.68
 te
 ms
 en
 gi
 ne
 er
 in
 g
Dep   8.25     9.60   1.35  11.91    12.23   0.32  16.61    17.88   1.27    2.94
 lo
 ym
 en
 t
 pl
 an
 ni
 ng
Pro   7.56    29.65  22.09  26.89    28.43   1.54  33.47    60.05  26.58   50.21
 gr
 am
 ma
 na
 ge
 me
 nt
 /
 su
 pp
 or
 t
Sys  29.33    69.73  40.40  50.76   102.87  52.11  83.71   140.51  56.81  149.31
 tem
 te
 st
 an
 d
 ev
 al
 ua
 ti
 on
Sen  64.89    87.64  22.74  53.93    53.57      -  30.28    18.38      -   10.48
 sor                                         0.36                  11.90
 te
 ch
 no
 lo
 gy
Oth      0        0      0      0    17.40  17.40      0        0      0   17.40
 er
 in
 it
 ia
 ti
 ve
 s
Spe      0        0      0      0    18.00  18.00      0        0      0   18.00
 ci
 al
 in
 te
 re
 st
Mis  53.24    56.64   3.39  69.18    74.99   5.81  51.22    54.52   3.30   12.50
 si
 on
 co
 mm
 on
BMD      0        0      0  33.48    31.81      -  32.22    28.58      -   -5.32
 O                                           1.68                   3.64
 Ma
 na
 ge
 me
 nt
Res      0    14.90  14.90      0    22.02  22.02      0    43.21  43.21   80.14
 ci
 ss
 io
 ns
 an
 d
 re
 du
 ct
 io
 ns
================================================================================
Tot  $370.  $745.62  $375.  $508.  $833.44  $325.  $504.  $978.09  $474.  $1,174
 al     62              00     44              00     09              00     .00
--------------------------------------------------------------------------------
Note:  Totals may not add due to rounding. 




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


*** End of document. ***