Defense Acquisitions: Improved Management Practices Could Help
Minimize Cost Growth in Navy Shipbuilding Programs (28-FEB-05,
GAO-05-183).
The U.S. Navy invests significantly to maintain technological
superiority of its warships. In 2005 alone, $7.6 billion was
devoted to new ship construction in six ship classes--96 percent
of which was allocated to four classes: Arleigh Burke class
destroyer, Nimitz class aircraft carrier, San Antonio class
amphibious transport dock ship, and the Virginia class submarine.
Cost growth in the Navy's shipbuilding programs has been a
long-standing problem. Over the past few years, the Navy has used
"prior year completion" funding--additional appropriations for
ships already under contract--to pay for cost overruns. This
report (1) estimates the current and projected cost growth on
construction contracts for eight case study ships, (2) breaks
down and examines the components of the cost growth, and (3)
identifies any funding and management practices that contributed
to cost growth.
-------------------------Indexing Terms-------------------------
REPORTNUM: GAO-05-183
ACCNO: A18416
TITLE: Defense Acquisitions: Improved Management Practices Could
Help Minimize Cost Growth in Navy Shipbuilding Programs
DATE: 02/28/2005
SUBJECT: Appropriated funds
Construction contracts
Construction costs
Cost analysis
Defense budgets
Funds management
Future budget projections
Labor costs
Military cost control
Military vessels
Naval procurement
Overhead costs
Past fiscal year
Ships
Cost estimates
Cost growth
Arleigh-Burke Class Destroyers
CVN-76 Aircraft Carrier
CVN-77 Aircraft Carrier
DDG-91 Destroyer
DDG-92 Destroyer
LPD-17 Amphibious Ship
LPD-18 Amphibious Ship
Nimitz Class Aircraft Carrier
San Antonio Class Amphibious Transport
Dock Ship
SSN 774 Submarine
SSN 775 Submarine
Virginia Class Submarine
******************************************************************
** This file contains an ASCII representation of the text of a **
** GAO Product. **
** **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced. Tables are included, but **
** may not resemble those in the printed version. **
** **
** Please see the PDF (Portable Document Format) file, when **
** available, for a complete electronic file of the printed **
** document's contents. **
** **
******************************************************************
GAO-05-183
United States Government Accountability Office
GAO Report to the Chairman, Subcommittee on Defense,
Committee on Appropriations, House of Representatives
February 2005
DEFENSE ACQUISITIONS
Improved Management Practices Could Help Minimize Cost Growth in Navy
Shipbuilding Programs
a
GAO-05-183
[IMG]
February 2005
DEFENSE ACQUISITIONS
Improved Management Practices Could Help Minimize Cost Growth in Navy
Shipbuilding Programs
What GAO Found
For the eight ships GAO assessed, the Congress has appropriated funds to
cover the $2.1 billion increase in the ships' budgets. The GAO's analysis
indicates that total cost growth on these ships could reach $3.1 billion
or even more if shipyards do not maintain current efficiency and meet
schedules. Cost growth for the CVN 77 aircraft carrier and the San Antonio
lead ship (LPD 17) has been particularly pronounced.
Increases in labor hour and material costs together account for 77 percent
of the cost growth on the eight ships. Shipbuilders frequently cited
design modifications, the need for additional and more costly materials,
and changes in employee pay and benefits as the key causes of this growth.
For example, the San Antonio's lead ship's systems design continued to
evolve even as construction began, which required rebuilding of completed
areas to accommodate the design changes. Materials costs were often
underbudgeted, as was the case with the Virginia class submarines and
Nimitz class aircraft carriers. For the CVN 77 carrier, the shipbuilder is
estimating a substantial increase in material costs.
Components of Cost Growth 5%
Overhead rate and labor rate increases
Material increases
Labor hour increases
Percentage of overall cost growth due to Percentage of overall cost growth
due to cost of shipbuilder construction costs Navy-furnished equipment
Source: Shipbuilder and Navy (data); GAO (analysis).
Navy practices for estimating costs, contracting, and budgeting for ships
have resulted in unrealistic funding of programs, increasing the
likelihood of cost growth. Despite inherent uncertainties in the ship
acquisition process, the Navy does not account for the probability of cost
growth when estimating costs. Moreover, the Navy did not conduct an
independent cost estimate for carriers or when substantial changes
occurred in a ship class, which could have provided decision makers with
additional knowledge about a program's potential costs. In addition,
contract prices were negotiated and budgets established without sufficient
design knowledge and construction knowledge. When unexpected events did
occur, the incomplete and untimely reporting on program progress delayed
the identification of problems and the Navy's ability to correct them.
United States Government Accountability Office
Contents
Letter
Results in Brief
Background
New Ships Continue to Cost More Than Budgeted
Labor and Materials Drive Increases in Construction Costs
Navy Funding and Management Practices Result in Insufficient
Provision for Risk
Conclusions
Recommendations
Agency Comments and Our Evaluation
1
2 3 7 10
19 28 29 30
Appendix I Scope and Methodology
Appendix II Arleigh Burke Class Destroyer 37
Program Description 37
Cost Experience on DDG 91 and DDG 92 39
Main Drivers of Cost Growth for DDG 91 DDG 92 39
Appendix III Nimitz Class Aircraft Carrier 45
Program Description 45
Cost Experience on CVN 76 and CVN 77 46
Main Drivers of Cost Growth for CVN 76 and CVN 77 48
Appendix IV San Antonio Class Amphibious Transport Dock Ship 55
Program Description 55
Cost Experience on LPD 17 and LPD 18 56
Main Drivers of Cost Growth for LPD 17 and LPD 18 58
Appendix V Virginia Class Submarine 62
Program Description 62
Cost Experience on SSN 774 and SSN 775 63
Main Drivers of Cost Growth for SSN 774 and SSN 775 65
Appendix VI GAO's Forecast of Additional Costs to Complete
Construction Contracts 73
Appendix VII Comments from the Department of Defense 75
Appendix VIII GAO Contacts and Staff Acknowledgments
Tables
Table 1: Overview of Navy Shipbuilding Programs Represented in
GAO's Case Studies 4 Table 2: Shipbuilding Budget Cost Categories 6 Table
3: Growth in Program Budgets for Case Study Ships 8 Table 4: GAO's
Forecasts of Additional Cost Growth
for Construction 9 Table 5: Growth in Labor Hour Costs 12 Table 6: Reasons
Given by Shipbuilders for Labor Hours Cost
Growth 13 Table 7: Growth in Material Costs 14 Table 8: Reasons Given by
Shipbuilders for Material Cost Growth 16 Table 9: Growth in Overhead Costs
and Labor Rates 17 Table 10: Reasons Given by Shipbuilders for Overhead
and Labor
Rate Cost Growth 18 Table 11: Target Prices for Case Study Ships 23 Table
12: Characteristics of Case Study Ships 34 Table 13: Major Events in the
Acquisition of DDG 91 and DDG 92 38 Table 14: Growth in Program Budgets
for Case Study Ships 39 Table 15: Growth in Labor Hour Costs 41 Table 16:
Growth in Overhead Costs and Labor Rates 42 Table 17: Growth in Material
Costs 43 Table 18: Major Events in the Acquisition of CVN 76 and CVN 77 46
Table 19: Growth in Program Budgets for Case Study Ships 47 Table 20:
GAO's Forecasts of Additional Cost Growth for
Construction 48 Table 21: Growth in Material Costs 49 Table 22: Growth in
Labor Hour Costs 51 Table 23: Historical Man-hours Used to Produce Prior
Ships
Compared to CVN 76 Negotiated Man-hours 52 Table 24: Growth in Overhead
Costs and Labor Rates 54 Table 25: Major Events in the Acquisition of LPD
17 and LPD 18 56 Table 26: Growth in Program Budgets for Case Study Ships
57
Table 27: GAO's Forecasts of Additional Cost Growth for
Construction 57 Table 28: Growth in Material Costs 59 Table 29: Growth in
Labor Hour Costs 59 Table 30: Growth in Overhead Costs and Labor Rates 60
Table 31: Major Events in the Acquisition of SSN 774 and SSN 775 63 Table
32: Growth in Program Budgets for Case Study Ships 64 Table 33: GAO's
Forecasts of Additional Cost Growth for
Construction 65 Table 34: Growth in Material Costs 67 Table 35: Growth in
Labor Hour Costs 69 Table 36: Growth in Overhead Costs and Labor Rates 71
Figures
Figure 1: Typical Production Times for Various Weapon Systems 5 Figure 2:
Components of Cost Growth 11 Figure 3: GAO and Shipbuilder Construction
Cost Growth
Forecasts for Case Study Ships 28 Figure 4: Arleigh Burke Class Destroyer
38 Figure 5: Average Sources of Cost Growth on DDG 91 and DDG 92 40 Figure
6: Nimitz Class Aircraft Carrier 45 Figure 7: Average Sources of Cost
Growth on CVN 76 and CVN 77 49 Figure 8: San Antonio Class Amphibious
Transport Dock Ship 55 Figure 9: Average Sources of Cost Growth on LPD 17
and LPD 18 58 Figure 10: Virginia Class Submarine 62 Figure 11: Average
Sources of Cost Growth on SSN 774 and
SSN 775 66 Figure 12: SSN 774 Lead Ship Labor Hour Growth 70 Figure 13:
Comparison of Shipbuilders' and GAO's Forecasts of
Additional Construction Costs for Six Classes of Ships
Actively under Construction 73
This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.
United States Government Accountability Office Washington, DC 20548
February 28, 2005
The Honorable C. W. Bill Young
Chairman, Subcommittee on Defense
Committee on Appropriations
House of Representatives
Dear Mr. Chairman:
U.S. Navy warships are the most technologically advanced in the world.
The United States invests significantly to maintain this advantage. In
2005
alone, the Navy devoted $7.6 billion to new ship construction in six ship
classes-96 percent of which was allocated to four classes: Arleigh Burke
class destroyer, Nimitz class aircraft carrier, San Antonio class
amphibious
transport dock ship, and the Virginia class submarine.
Cost growth in the Navy's shipbuilding programs has been a long-standing
problem-one that the Congress has identified and responded to by
providing both additional funding and direction to the Navy. Over the past
few years, the Navy has used "prior year completion" funding-additional
appropriations for ships already under contract-to pay for cost overruns.
Because of the size and routine occurrence of prior year funding, we were
asked to analyze cost overruns on Navy shipbuilding programs.
Specifically, this report (1) estimates the current and projected cost
growth on selected ship construction contracts, (2) breaks down and
examines the components of the cost growth, and (3) identifies any
funding and management practices that contribute to cost growth.
To address these objectives, we looked at cost growth in the four classes
of ships that account for the majority of the funding for new shipbuilding
and prior year bills, focusing on ships with construction contracts that
were more than 30 percent complete at the time we began our review.
Within each class, we selected two ships currently under contract as case
studies: DDG 91 and DDG 92 in the Arleigh Burke class of destroyers,
CVN 76 and CVN 77 in the Nimitz class of aircraft carriers, LPD 17 and
LPD 18 in the San Antonio class of transports, and SSN 774 and SSN 775 in
the Virginia class of submarines. To estimate the total projected cost
growth on construction contracts, we used contractor performance
reports, projecting high and low estimates for the costs to complete the
ships in the four classes we reviewed. We looked at cost growth by
comparing the initial budget request to the Congress and the updated
Results in Brief
budget included in the 2005 President's budget and by comparing the
initial contract award to the latest estimate at completion. The latest
estimate at completion includes changes to the original baseline or scope
of work. For all ships currently under construction, we also estimated
total cost growth since contract award, using contractor performance
reports to project high and low estimates for the costs to complete
construction of these ships. To break down and examine the components of
cost growth on the eight case study ships, we analyzed the Navy's cost
estimates, its budget requests to the Congress, contractor performance
reports, and other cost data for each of the eight case study ships. To
assess funding and management practices, we spoke with the shipbuilders,
Navy and Defense Contract Audit Agency officials, and reviewed supporting
documentation. Our work was conducted between July 2003 and December 2004
in accordance with generally accepted government auditing standards. Our
analyses and forecasts were based on data available to us in July 2004. To
the extent significant changes occurred, we incorporated information from
the fiscal year 2006 President's budget. For a complete description of our
scope and methodology, see appendix I. Details on the eight case study
ships are discussed in appendixes II to V.
The Navy's shipbuilding programs continue to experience significant cost
growth. For the eight case study ships alone, the Congress has
appropriated funds to cover a $2.1 billion increase in the ships' budgets.
Cost growth was pronounced for the CVN 77 carrier and for the lead ships
in the two new classes we looked at-the Virginia class and especially the
San Antonio class. We estimated cost growth could exceed $3 billion, and
these estimates are likely understated because they assume that the
shipyards will maintain their current efficiencies and meet scheduled
milestones. Thus, additional appropriations, in excess of $1 billion, will
be needed to cover the additional cost growth.
Increases in labor hour and material costs account for 78 percent of the
cost growth on the eight ships we reviewed, while overhead and labor rate
increases account for 17 percent. Navy-furnished equipment-including
radars and weapon systems-represent just 5 percent of the cost growth.
Shipbuilders cited a number of direct causes for the labor hour, material,
and overhead cost growth in the eight ships. The most common causes were
related to design modifications, the need for additional and more costly
materials, and changes in employee pay and benefits. For example, the lack
of design maturity when introducing new technologies led to rework,
increasing growth in labor hours for most of the ships. The design of ship
systems for LPD 17 continued to evolve even as construction
proceeded. As a result, workers were required to rebuild completed areas
of the ship to accommodate design changes. Growth in materials costs was
due, in part, to the Navy's and shipbuilders' underbudgeting of these
costs. For example, the materials' budget for the first four Virginia
class submarines was $132 million less than quotes received from vendors
and subcontractors at contract award. Price increases also contributed to
the growth in materials costs.
Navy practices for estimating costs, contracting, and budgeting for ships
have resulted in unrealistic funding of programs, increasing the
likelihood of cost growth. Despite inherent uncertainties in the ship
acquisition process, the Navy does not measure or provide for the
probability of cost growth when estimating costs. Moreover, the Navy did
not conduct independent cost estimates for carriers, which could have
provided decision makers with additional knowledge about a program's
potential costs. In addition, contract prices were negotiated and budgets
established without making full use of design knowledge and construction
experience. Finally, when unexpected events occurred, the incomplete and
untimely reporting on program progress delayed the identification of
problems and the Navy's ability to correct them.
We are making seven recommendations aimed at improving the Navy's
processes for developing cost estimates, establishing realistic contract
prices and ship budgets, and providing timely and complete reporting on
program costs to alert managers to potential problems. In its comments on
a draft of this report, DOD concurred with two of our recommendations and
partially concurred with five. We believe the Navy needs to take concrete
action to establish realistic estimates, prices, and budgets and to
improve the quality of cost reporting.
Background The U.S. Navy currently operates 288 surface ships and
submarines. Four ship classes, with 23 ships under construction or
recently completed, make up 96 percent of the Navy's fiscal year 2005
budget for new construction shipbuilding. (See table 1.)
Table 1: Overview of Navy Shipbuilding Programs Represented in GAO's Case
Studies
Percent of Navy's
fiscal year 2005 new
construction and Ships under
prior year construction or
Mission GAO's case
Ship class shipbuilding budgeta recently completed study shipsb
Arleigh Burke Destroyers provide 46% 13 o DDG 91 (follow-on
destroyer ship)
offensive and defensive o DDG 92 (follow-on
capabilities; can ship)
operate independently
or as part of strike
groups
Nimitz aircraft Nuclear-powered N/Ac 2 o CVN 76 (follow-on
carrier ship)
aircraft carriers form o CVN 77 (follow-on
building block of Navy's ship)
forward-deployed
peacetime presence,
crisis response, and
warfighting forces
San Antonio These amphibious 14% 5 o LPD 17 (lead ship)
amphibious
transport dock ship ships provide a sea-based o LPD 18 (follow-on
platform for ship)
transporting,
embarking, and landing
Marines and their
equipment and
supplies during an
assault
Virginia class submarine This new class of nuclear submarines is designed
to combat enemy submarines and surface ships; fire cruise missiles at land
targets; and provide improved surveillance and special operations support
36% 4 o SSN 774 (lead ship)
o SSN 775 (lead ship)
Sources: Navy (data); GAO (presentation).
aIncluding completion of ships authorized in prior years
bA lead ship is the first to be built in a class or the first to be built
after a major redesign of a class of ships. If two different shipbuilders
are constructing ships that fall within the same class, their first ships
are also referred to as lead ships. Follow-on ships are those built after
the lead ship.
cNot applicable. CVN 76 and CVN 77 were funded in earlier fiscal years and
are not included in these percentages.
Navy ships are complex defense systems, using advanced designs with
state-of-the-art weapons, communications, and navigation technologies.
Ships require many years to plan, budget, design, and build. Like other
weapon acquisition programs, ship acquisitions begin with developing a
system design. For ships, system design is followed by a detail design
phase where specific construction plans are developed. Ship construction
follows and typically takes 4 to 7 years. Construction time for other
defense systems is much shorter-a fighter aircraft takes about 2 years
from start of production to roll out from the factory floor; a tank takes
about a year. (See fig. 1.)
Figure 1: Typical Production Times for Various Weapon Systems
Type of weapon system
Aircraft carrier
Submarine
Destroyer
Aircraft
Tank
01234567
Years in production
Source: Navy (data); GAO (presentation).
Note: A varying number of years of preliminary planning and early design
work precede the start of production.
The long construction times increase the uncertainty that ship cost
estimates-and budgets-must provide for. Moreover, the total cost for a
ship must be budgeted for in its first year of construction. Provisions
are made in the event cost growth occurs during construction. The Navy's
budgeting for cost growth has changed over the past 2 decades. During the
early 1970s and through most of the 1980s, the Navy used program cost
reserves built into ship construction budgets and the Ship Cost Adjustment
process to manage cost growth. During the 1980s, the Navy procured an
average of 17 ships each year. In fiscal year 1988, the Navy removed
program cost reserves from ship construction budgets and began exclusively
using the Ship Cost Adjustment process, shifting funding between
shipbuilding construction programs underrunning cost to programs that were
overrunning costs. Following the end of the Cold War, the Navy decreased
the procurement rate of ships to about 6 per year. Beginning in fiscal
year 1999, cost increases could no longer be covered
using the Ship Cost Adjustment process because no shipbuilding program was
under cost. In 2001, the process was eliminated, which required the Navy
to fund cost growth through the current mechanism of prior year completion
bills.
Components of The cost of building a ship can be broken down into four
main
Shipbuilding Costs components: labor, material, and overhead associated
with the shipbuilders' contract for the basic ship, and Navy-furnished
equipment- that is, items purchased by the Navy and provided to the
contractor for installation on the ship. (See table 2.) The shipbuilding
contract also includes profit (referred to as fee).1
Table 2: Shipbuilding Budget Cost Categories Construction
C
Labor Materials a Overheada Navy-furnished
equipment
o Metals (steel, copper, o
Labor hours for Medical insurance o Items
production, purchased by the Navy and
engineering and titanium) o provided to the
other direct Pensions shipbuilder for
installation on the
support o Tools ship; items
osts based on labor hours and o Miscellaneous parts (pipe, o Holiday
pay include ship weapon systems,
o Facilities maintenance propulsion equipment, and
the labor rate (the hourly wage cables) and utilities electronicspaid to
workers) o Subcontracts o Taxes
Sources: Shipbuilder and Department of Defense (data); GAO (analysis).
aList is not exhaustive.
Types of Shipbuilding Contracts
Two broad categories of contracts are used to procure ships: fixed-price
and cost-reimbursement. Fixed price contracts provide for a firm price or
an adjustable price with a ceiling price, a target price, or both. If the
ceiling is reached the shipbuilder is generally responsible for all
additional costs. Cost reimbursement contracts provide for payment of
allowable incurred costs, to the extent prescribed in the contract. If the
ship cannot be completed within agreed upon cost limits, the government is
responsible for the additional costs to complete.2
1 The contract also includes funds for the cost of money. 2 See Federal
Acquisition Regulation Part 16.
The level of knowledge, or certainty, in the cost estimates for a ship is
key to determining which type of contract to use. Contracts for the first
ship of a new class are often negotiated as cost-reimbursable contracts
because these ships tend to involve a high-level of uncertainty and, thus,
high cost risks. Cost reimbursement contracts were used to procure the San
Antonio and Virginia class ships we reviewed. More mature shipbuilding
programs, where there is greater certainty about costs, are typically
fixed-price contracts with an incentive fee (profit). Fixed-price
contracts were used to procure the Arleigh Burke and Nimitz class ships we
reviewed. Both costreimbursable and fixed-price incentive fee contracts
can include a target cost, a target profit, and a formula that allows the
profit to be adjusted by comparing the actual cost to the target cost.
Construction contracts for ships generally include provisions for
controlling cost growth with incentive fees, whereby the Navy and the
shipbuilder split any savings when the contract cost is less than its
anticipated target. Conversely, when costs exceed the target, the excess
is shared between the Navy and the shipbuilder.
New Ships Continue to Cost More Than Budgeted
Ship cost growth continues to pose additional funding demands on the
budget. Budgets for the eight case study ships alone have required
increases of $2.1 billion, and Congress has appropriated funds to cover
these increases. However, the total projected cost growth on contracts for
the eight ships is likely to be higher. Consequently, the Navy will need
in excess of $1 billion in additional appropriations to cover the total
projected cost growth. Cost growth was more pronounced for the lead ships
in the two new classes we looked at-the Virginia class and especially the
San Antonio class-than the more mature Arleigh Burke and Nimitz classes.
(Our forecasts for cost growth on all ships that are more than 30 percent
complete are shown in appendix VI.)
The fiscal year 2005 budget for the eight case study ships was about $20.6
billion-representing cost growth of $2.1 billion above the initial budget
request of $18.5 billion for these ships. (See table 3.) Ship construction
costs comprise the majority of this increase.
Table 3: Growth in Program Budgets for Case Study Ships
Dollars in millions
Initial and fiscal year 2005 President's budget Difference in budgets
Difference due to Navy-furnished equipment Difference due to construction
costsc
Case study ship Initiala FY2005b Total difference
DDG 91 $917 $997 $80 $43 $37
DDG 92 925 979 55 (7)d
CVN 76 4,476 4,600 124 (128)e
CVN 77 4,975 5,024 49 100 (51)f
LPD 17 954 1,758 804 21
LPD 18 762 1,011 249 3
SSN 774 3,260 3,682 422 95
SSN 775 2,192 2,504 312 18
Total 18,461 20,556 2,095 145 1,951
Sources: Navy (data); GAO (presentation).
aEstimated cost from the President's budget submission for year of ship
authorization.
bIncludes all prior year requests through fiscal year 2005.
cPart of increased cost is due to changes in the scope of the contract.
dNegative reflects savings resulting from the use of a more economical
warfare system than was initially budgeted on the DDG 92.
eNegative reflects savings garnered from Navy-furnished reactor plant
equipment.
fNegative reflects shifting of funds from the construction contract to
Navy-furnished equipment.
We were not able to determine how much of this increase was due to changes
in the scope of the contract and how much of the growth funded increases
in the costs of completing the initial contract scope. Amounts identified
by shipbuilders and Navy program offices differed substantially. However,
the initial program budgets included funding to support changes in the
scope of the construction contract. These funds amounted to a small share
of the initial program budget: 3 percent for DDGs 91 and 92; 5 percent for
CVN 76 and CVN 77; 7 percent for LPD 17 and 4 percent for LPD 18; and 3
and 4 percent for SSNs 774 and 775, respectively.
While the Congress has appropriated funds to cover a $2.1 billion increase
in the ships' costs, more funds will likely be needed to cover additional
cost growth likely for these eight ships. At the time we completed our
analysis in 2004, we calculated a range of the potential growth for the
eight
case study ships and found that the total projected cost growth would
likely exceed $2.8 billion and could reach $3.1 billion. (See table 4.)
Table 4: GAO's Forecasts of Additional Cost Growth for Construction
Dollars in millions
Forecasts based on data available July 2004
Amount already GAO's
requested forecast of
Percent of ship to cover additional GAO's forecast
contractor's cost of
Case study construction increased cost growth total cost
ship completed growtha
DDG 91 Delivered $37 $37-37
DDG 92 Delivered 62 62-62
CVN 76 Delivered 252 252-252
CVN 77 45 (51)b $485-637 434-586c
LPD 17 93 784 112-197 896-981
LPD 18 69 246 102-136 348-382
SSN 774 Delivered 327 327-327d
SSN 775 88 294 103-219 397-513
Total
growth 1,951 802-1,189 $2,753-3,140
Sources: Shipbuilder and Navy (data); GAO (analysis).
aForecast reflects expected price to the Navy.
bNegative reflects shifting of funds from the construction contract to
Navy-furnished equipment.
cThe 2006 budget submission indicates $908 million additional cost growth
on CVN 77 above the fiscal year 2005 budget.
dThe Navy has requested an additional funding to cover completion of SSN
774.
These cost growth estimates have already proven to be too conservative. In
its fiscal year 2006 budget submission, the Navy recognizes an additional
cost growth of $223 million for SSN 775 and $908 million for CVN 77 above
its fiscal year 2005 request. In addition, our estimates assumed that the
shipyards will maintain their current efficiency through the end of their
contracts and meet scheduled milestones. Any slips in efficiency and
schedules would likely result in added costs. For example, the delivery
date for SSN 775 is expected to slip by as many as 9 months, which,
according to the fiscal year 2006 President's budget has increased the
final cost of the ship even more. According to program officials, the
delivery date for the LPD 17 has been changed from December 2004 to May
2005, and the delivery date for the CVN 77 is expected to slip into 2009.
Labor and Materials Drive Increases in Construction Costs
Cost growth on new ships has a number of implications. Most tangible,
perhaps, is the significant portion of the ship construction budget that
must be devoted to overruns on ships already under construction. From
fiscal years 2001 to 2005, 5 to 14 percent of the Navy's ship construction
budget, which totaled about $52 billion over the 5-year period, went to
pay for cost growth for ships funded in prior years. This reduces the
buying power of the budget for current construction and can slow the pace
of modernization. The Navy is in the early stages of buying a number of
advanced ships, including the Virginia class submarine, DD(X) destroyer,
CVN 21 aircraft carrier, and Littoral Combat Ship. The Navy's ability to
buy these ships as scheduled will depend on its ability to control cost
growth.
Increases in labor hour and material costs account for 78 percent of the
cost growth on shipbuilding construction contracts, while overhead and
labor rate increases account for 17 percent. Navy-furnished equipment3-
including radars and weapon systems-represents just 5 percent of the cost
growth. (See fig. 2.) Shipbuilders cited a number of direct causes for the
labor hour, material, and overhead cost growth in the eight case study
ships. The most common causes were related to design modifications, the
need for additional and more costly materials, and changes in employee pay
and benefits.
3 These costs are items provided by the Navy to the contractor for
installation on the ship. The Navy pays for this equipment-not the
shipbuilder.
Figure 2: Components of Cost Growth
Labor hour increases for the eight case study ships ranged from 33 percent
to 105 percent-for a total of 34 million extra labor hours. For example,
the shipbuilders for LPD 17 and CVN 76 each needed 8 million additional
labor hours to construct the ships
Cost growth due to increased labor hours totaled more than $1.3 billion.
(See table 5.) While the total dollars were the greatest for LPD 17 ($284
million), the labor cost as a percent of total cost growth was the
greatest for DDG 91 (105 percent).
Overhead rate and labor rate increases
Material increases
Labor hour increases
Percentage of overall cost growth due to shipbuilder construction costs
Percentage of overall cost growth due to cost of Navy-furnished equipment
Source: Shipbuilder and Navy (data); GAO (analysis).
Note: Total growth in construction costs is $3.2 billion, based on shipbuilders'
estimate at completion.
Design Changes and Lack of Skilled Labor Contributed to Labor Hour Cost
Growth
Table 5: Growth in Labor Hour Costs
Dollars in millions
Analysis based on data available July 2004
Overhead and Labor hour
cost as a
Shipbuilder labor rate costs Total cost due percent of
reported on to total
Case study labor cost growth increased labor increased labor contract
ship hours hours growth
DDG 91 $23 $24 $47 105%
DDG 92 43 42 85
CVN 76a 78 144 222
CVN 77a 75 107 182
LPD 17b 182 102 284
LPD 18 117 67 184
SSN 774 149 10 159
SSN 775 218 (38) 180
Total $885 $458 $1,342
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures all costs associated with
labor hour growth-including overhead and labor rates. Methodology is
discussed in appendix 1.
aContractor performance reports included $63 million in overhead costs for
CVN 76 and $40 million for CVN 77 that have been disallowed (not charged
to the government).
bLPD 17 relied heavily on subcontracts with partners (Bath Iron Works and
Raytheon) to design the ship. Since these costs are captured as material,
we did not include them in our analysis of labor cost increases.
The lack of design and technology maturity led to rework, increasing the
number of labor hours for most of the case study ships. For example, the
design of LPD 17 continued to evolve even as construction proceeded. When
construction began on DDG 91 and DDG 92-the first ships to incorporate the
remote mine hunting system-the technology was still being developed. As a
result, workers were required to rebuild completed areas of the ship to
accommodate design changes. Most of the shipbuilders cited a lack of
skilled workers as a driver behind labor hour cost growth. According to
the shipbuilders we interviewed, many of the tasks needed to build ships
are complex and require experienced journeymen to efficiently carry them
out. Yet, the majority of the shipbuilders noted that the shipyards have
lost a significant portion of their highly skilled and experienced
workers. Delays in delivery of materials also resulted in increased labor
hours. Table 6 shows the reasons for labor hour increases for each case
study ship.
Table 6: Reasons Given by Shipbuilders for Labor Hours Cost Growth
Case study ship Reasons for increase
DDG 91 o Inexperienced laborers
o Design upgrades that result in rework
DDG 92 o Introduction of a new construction facility, setting workers
back on the learning curve
o Design upgrades that result in rework and workarounds
o Strike increased number of hours needed to construct ship
CVN 76 o Less-skilled workers due to demands for labor on other
programs at shipyard
o Extensive use of overtime
o Design changes resulting in rework
CVN 77 o Late material delivery results in delays and workarounds
o Design changes resulting in rework
LPD 17 o Inexperienced subcontracted labor
o Design difficulties led to doing work out of sequence and rework
o Schedule delays
o Bused workers to meet labor shortages
LPD 18 o Increases in LPD 17 translated into more hours for LPD 18
SSN 774 o Late material delivery
o First in class design issues
SSN 775 o Quality problems and design changes
o Inclusion of non-recurring labor hours
Sources: Shipbuilder (data); GAO (analysis).
Underbudgeting and Price For several of the case study ships, the costs of
materials increased Increases Contributed to dramatically above what the
shipbuilder had initially planned. (See Materials Cost Growth table 7.)
Materials cost was the most significant component of cost growth
for three ships: LPD 17, SSN 775, and CVN 76. However, for LPD 17, which
experienced over 100-percent growth in material costs, 70 percent of the
material cost increases were actually costs for subcontracts to support
design of the lead ship.
Table 7: Growth in Material Costs
Dollars in millions
Analysis based on data available July 2004
Total dollars due to Material cost as a
percent of
Case study increased material Percent increase total contract growth
ship costs
DDG 91 ($22) (13%) (49%)
DDG 92 30 20
CVN 76 294 43
CVN 77 134 13
LPD 17 400 103
LPD 18 93 39
SSN 774 141 43
SSN 775 209 56
Total $1,280 38%
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events.
Growth in materials costs was due, in part, to the Navy and shipbuilders'
underbudgeting for these costs. For example, the materials budget for the
first four Virginia-class submarines was $132 million less than quotes
received from vendors and subcontractors. The shipbuilder agreed to take
on the challenge of achieving lower costs in exchange for providing in the
contract that the shipbuilder would be reimbursed for cost growth in high
value, specialized materials. In addition, the materials budget for CVN 76
and CVN 77 was based on an incomplete list of materials needed to
construct the ship, leading to especially sharp increases in estimated
materials costs. In this case, the Defense Contract Audit Agency
criticized the shipbuilder's estimating system, particularly the system
for material and subcontract costs, and stated that the resulting
estimates "do not provide an acceptable basis for negotiation of a fair
and reasonable price." Underbudgeting of materials has contributed to cost
growth recognized in the fiscal year 2006 budget.
Price increases also contributed to the growth in materials costs. For
example, the price of array equipment on the Virginia class submarines
rose by $33 million above the original price estimate. In addition to
inflation, a limited supplier base for highly specialized and unique
materials made ship materials susceptible to price increases.4 According
to the shipbuilders, the low rate of ship production has affected the
stability of the supplier base-some businesses have closed or merged,
leading to reduced competition for the services they once produced and
that may be a cause of higher prices. In some cases, the Navy lost its
position as a preferred customer and the shipbuilder had to wait longer to
receive materials. With a declining number of suppliers, more ship
materials contracts have gone to single and sole source vendors. Over 75
percent of the materials for the Virginia class submarines-which were
reduced in number from 14 to 9 ships over a 10-year period-is produced by
single source vendors.
Spending on subcontracts and leased labor also increased material costs on
some case study ships.5 On LPD 17, for example, subcontracts to support
lead ship design accounted for 70 percent of the increase in material
costs. Table 8 highlights the various reasons cited for increased
materials costs on case study ships.
4 Cost estimates are based, in part, on the number of units produced and
learning curves- the more units produced, the less expensive each unit is
expected to be. Thus, if contractors and subcontractors are assured a
high, consistent level of business, they are able to produce the ship and
ship parts at a lower cost. Conversely, if purchases are erratic or dip to
historically low levels, the ship and ship parts will be more expensive to
produce, although the exact amount is uncertain.
5 Subcontracted labor is labor performed to a fixed price contract. Leased
labor is the employment of outside workers under the direct supervision of
the shipyard management and foreman systems.
Table 8: Reasons Given by Shipbuilders for Material Cost Growth
Case study ship Reasons for growth
DDG 91 o Consolidation with Northrop Grumman allowed for quantity
material buy savings
DDG 92 o Rework requiring additional tools, utilities, and shop stock
o Information technology costs shifted from overhead to materials
CVN 76 o Increases in costs for specialized materials
o Underbudgeted material costs
o Accounting changes
o Additional subcontracting
CVN 77 o Increases in costs for specialized materials
o Underbudgeted material costs
LPD 17 o Subcontractor engineering design efforts
o Design tool development, originally assumed to be funded by the state
resulted in additional costs to Northrop Grumman.
LPD 18 o Increases in LPD 17 translated into more costs for LPD 18
SSN 774 o Lack of suppliers for highly unique materials
o Immature design on material components
SSN 775 o Lack of suppliers for highly unique materials
o Nonrecurring costs for computer integration
Sources: Shipbuilder (data); GAO (analysis).
Program Overhead and Program overhead costs, which include increases in
labor rates,
Labor Rates Account for represented approximately 17 percent of the total
cost growth for the eight
Remaining Ship case study ships. (See table 9.) While increases in
overhead dollars totaled
Construction Cost more than $1 billion, almost half of the increase was
related to growth in
Increases labor hours. (See table 9.)
Table 9: Growth in Overhead Costs and Labor Rates
Dollars in millions
Analysis based on data available July 2004
Increase in Increase in overhead Overhead cost
overhead as a
Shipbuilder related to related to overhead percent of
reported growth in and total
Case study overhead growth labor hours labor rates contract
ship growth
DDG 91 $43 $24 $20 44%
DDG 92 56 42 14
CVN 76a 263 144 119
CVN 77a 219 107 113
LPD 17 277 102 175
LPD 18 177 67 110
SSN 774 0 10 (10) (3)
SSN 775 0 (38) 38
Total $1,035 $457 $579
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures only costs associated with
overhead and labor rate changes. Increases in overhead related to growth
in labor hours are captured in the analysis of labor hour increases.
aContractor performance reports included $63 million in overhead costs for
CVN 76 and $40 million for CVN 77 that have been disallowed (not charged
to the government).
Increases in program overhead were largely due to decreased workload at
the shipyards. Six of the eight case study ships experienced increased
overhead because there were fewer programs to absorb shipyard operation
costs. Increases in benefit costs, such as pensions and medical care
costs, and labor rate increases-the result of negotiations with labor
unions and inflation-also drove up program overhead costs. Table 10
highlights the various reasons cited for increased overhead costs on case
study ships.
Table 10: Reasons Given by Shipbuilders for Overhead and Labor Rate Cost
Growth
Case study ship Reasons for growth
DDG 91 o Pension plans affected by financial market changes
o Increase in medical benefit costs
o Union negotiations increase labor rates
o Loss of workload
DDG 92 o Medical care cost increases due to inflation and loss of
favorable medical care contract
o Loss of workload
CVN 76 o Changes in accounting of overhead
o Union negotiations following strike increase labor rates
CVN 77 o Changes in accounting of overhead
o Union negotiations following strike increase labor rates
o Medical care cost increases
o Capital investments
o Pension plans affected by financial market changes
o Workload changes
LPD 17 o Pension plans affected by financial market changes
o Loss of anticipated workload
o An over 2-year delay in lead ship delivery and change in the
procurement schedule
LPD 18 o Pension plans affected by financial market changes
o Loss of anticipated workload
o An over 2-year delay in lead ship delivery and change in the
procurement schedule
SSN 774 o Changes in pension, health care, and workman's compensation
o Overhead rates decreased due to increased workload
SSN 775 o Loss of expected business and training new workers
o Additional costs to restart submarine production capability at the
shipyard
Sources: Shipbuilder (data); GAO (analysis).
Navy-Furnished Navy-furnished equipment covers the costs for the
technologies and
Equipment equipment items-such as ship weapon systems and electronics-
purchased by the Navy and provided to the contractor for installation on
the ship. While Navy-furnished equipment accounts for 29 percent of the
budget for the eight case study ships, such equipment accounted for only 6
percent of the total cost growth. According to Navy officials, much of the
Navy-furnished equipment is common among many programs and,
therefore, benefits from economies of scale. However, the integration and
installation of these systems-especially the warfare systems-contributes
to cost growth and is captured in the shipbuilders' costs rather than
Navyfurnished equipment.
There was considerable variance from program to program. In addition, in
some cases, decreases and increases in Navy-furnished equipment were the
result of funds being reallocated. For example, the Integrated Warfare
System on CVN 77 was originally funded through the shipbuilder
construction contract, but was later deleted from the contract in favor of
an existing system furnished by the Navy.
Navy practices for estimating costs and for contracting and budgeting for
ships have resulted in unrealistic funding of programs and when unexpected
events occur, tracking mechanisms are slow to pick them up. Tools exist to
manage the challenges inherent in shipbuilding, including measuring the
probability of cost growth when estimating costs, making full use of
design and construction knowledge to negotiate realistic target prices,
and tracking and providing timely reporting on program costs to alert
managers to potential problems. For the eight case study ships, however,
the Navy did not effectively employ them to mitigate risk.
Navy Funding and Management Practices Result in Insufficient Provision for
Risk
Navy Estimates Do Not Capture Uncertainty and Are Often Not Independently
Evaluated
In developing cost estimates for the eight case study ships, Navy cost
analysts did not conduct uncertainty analyses6 to measure the probability
of cost growth, nor were independent estimates conducted for some
ships-even in cases where major design changes had occurred. Uncertainty
analyses and independent estimates are particularly important given the
inherent uncertainties in the ship acquisition process, such as the
introduction of new technologies and volatile overhead rates over time,
creating a significant challenge for cost analysts to develop credible
initial cost estimates. The Navy must develop cost estimates as much as 10
years before ship construction begins-before many program details are
known. As a result, cost analysts have to make a number of assumptions
about certain ship parameters, such as weight, performance,
6 According to Navy cost analysts, while they did not conduct uncertainty
analyses, they did perform sensitivity analyses in which they examined the
effects of different variables, including changes in procurement
quantities and labor rates.
or software, and about market conditions, such as inflation rates,
workforce attrition, and supplier base.
In the eight case study ships we examined, cost analysts relied on the
actual cost of previously constructed ships without adequately accounting
for changes in the industrial base, ship design, or construction methods.
Cost data available to Navy cost analysts were based on higher ship
construction rates from the 1980s. As a result, these data were based on
lower costs due to economies of scale-which were not reflective of the
lower procurement rates after 1989. In addition, in developing cost
estimates for DDG 91, DDG 92, LPD 17, and SSN 774, cost analysts relied on
actual cost data from previous ships in the same class or a similar class
but that were less technologically advanced. By using data from less
complex ships, Navy cost analysts tended to underestimate the costs needed
to construct the ships.
For CVN 76, cost analysts used proposed costs from CVN 74 with adjustments
made for design changes and economic factors. However, CVN 74 and CVN 75
were more economical ships because both were procured in a single
year-which resulted in savings from economies of scale. While cost
analysts adjusted their estimates to account for the single-ship buy,
costs increased far beyond the adjustment. Even in more mature
programs-like the Arleigh Burke destroyers and the Nimitz aircraft
carriers-improved capabilities and modifications made the costs of
previous ships in the class essentially less analogous.
Other unknowns also led to uncertain estimates in the case study ships.
Labor hour and material costs were based not only on data from previous
ships but also on unproven efficiencies in ship construction. We found
analysts often factored in savings based on expected efficiencies that
never materialized. For example, cost analysts anticipated savings through
the implementation of computer-assisted design/computer-assisted
manufacturing for LPD 17, but the contractor had not made the requisite
research investments to achieve the proposed savings. Similar unproven or
unsupported efficiencies were estimated for DDG 92 and CVN 76. Changes in
the shipbuilders' supplier base also created uncertainties in the
shipbuilders' overhead costs.
Despite these uncertainties, the Navy did not test the validity of the
assumptions made by the cost analysts in estimating the construction
costs for the eight case study ships nor did the Navy identify a
confidence level for estimates.7 Specifically, it did not conduct
uncertainty analyses, which generate values for parameters that are less
than precisely known around a specific set of ranges. For example, if the
number of hours to integrate a component onto a ship is not precisely
known, analysts may put in a low and high value. The estimate will
generate costs for these variables along with other variables such
as-weight, experience, and degree of rework. The result is a range of
estimates that enables cost analysts to make better decisions on likely
costs. Instead, the Navy presented its cost estimates as unqualified point
estimates, suggesting an element of precision that cannot exist early on
and obscures the investment risk remaining for the programs. While
imprecision decreases during the program's life cycle as more information
becomes known about the program, experts emphasize that to be useful, each
cost estimate should include an indication of its degree of uncertainty,
possibly as an estimated range or qualified by some factor of confidence.
Other services qualify their cost estimates by determining a confidence
level of 50 percent.
The Navy also did not conduct independent cost estimates for some ships,
which is required at certain major acquisition milestones.8 Independent
cost estimates can provide decision makers with additional insight into a
program's potential costs-in part because these estimates frequently use
different methodologies and may be less burdened with organizational bias.
Independent cost analysts also tend to incorporate cost for risk as they
develop their estimates, which the Navy cost analysts did not do. As a
result, these independent estimates tend to be more conservative-
forecasting higher costs than those forecast by the program office.
Department of Defense officials considered the CVN 68 and DDG 51 programs
mature programs and, therefore, did not require independent estimates.
Yet, an independent cost estimate has never been conducted on a CVN 68
class carrier because the program for this class of ships began prior to
the establishment of an independent cost-estimating group in DOD. However,
Navy officials noted that every carrier is a new program,
7 A level of confidence depicts how much confidence the estimators have,
stated as a percentage, in a budget or schedule estimate. The higher the
confidence level, the lower the risk.
8 10 U.S.C. S:2434. These milestones include Milestone B, which marks the
beginning of the system development and demonstration phase. Milestone C
marks the beginning of the production and deployment phase.
different from previous carriers. Although an independent cost estimate
was conducted for the DDG 51 program, it was conducted in 1993, and since
that time, the DDG ships have undergone four major upgrades.
The Navy has begun taking some actions to improve its cost estimating
capabilities. For example, future programs will be funded at the DOD
independent estimators' level, which should provide a more conservative
estimate and include risk analysis. In addition, Navy officials told us
that they are in the process of revising cost estimating guidance to
include requirements for risk and uncertainty analysis. The degree to
which this guidance will enable the Navy to provide more realistic cost
estimates for its shipbuilding programs will depend on how it will be
implemented on individual programs.
Contract Prices Negotiated
and Budgets Set Without
Making Full Use of
Design Knowledge and
Construction Experience
Uncertainty about costs is especially high for new classes of ships, since
new classes incorporate new designs and new technologies. Yet, the Navy's
approach to negotiating contract target prices for construction of the
lead ship and early follow-on ships does not manage this uncertainty
sufficiently-evidenced by substantial increases in the prices of the first
several ships. Target prices for detail design and construction of the
lead and early follow-on ships are typically negotiated at one time.9 In
these cases the Navy does not make use of knowledge gained during detailed
design or during construction of the lead ship to establish more realistic
prices. When this approach to negotiating prices was used, it also
affected the information that was available to the Congress at the time it
funded construction of lead and follow-on ships.
Target prices for all of the case study ships increased, but, as shown in
table 11, the increase was greater for the two San Antonio class ships and
the two Virginia class ships-both new classes of ships. Increases in the
target prices of the LPD 17 and LPD 18 were particularly pronounced,
reaching 139 and 95 percent, respectively.
9 The Virginia class was an exception to this practice. For this program,
the Navy separated the funding of detail design from construction.
Table 11: Target Prices for Case Study Ships
Dollars in millions
Analysis based on data available July 2004
Shipbuilders'
Case study Initial target price estimated price Cost growth Percent change
ship
DDG 91 $355 $390 $35 10%
DDG 92 351 422 71
CVN 76 2,967 3,391 424
CVN 77 3,446 3,879 434
LPD 17 644 1,539 896
LPD 18 391 764 373
SSN 774 1,028 1,301 273
SSN 775 1,084 1,488 404
Total $10,266 $13,174 $2,910 28%
Sources: Shipbuilder and Navy (data); GAO (analysis).
The realism of target prices reflects the Navy's approach to negotiating
contract prices-the Navy negotiates target prices for the first several
ships at a stage of the program when uncertainty is high and knowledge
limited. For example, for the San Antonio class ships, the Navy negotiated
prices for the detail design and construction of the lead ship (LPD 17)
and the first two follow-on ships (LPD 18 and LPD 19) at the same time.10
By negotiating target prices for these ships before detail design even
began, target prices for these three ships did not benefit from
information gained during detail design about the materials and equipment
or specific processes that will be used to construct the ship. Target
prices for the follow-on ships, LPD 18 and LPD 19, did not benefit from
knowledge gained in initial construction of LPD 17. In contrast, for the
Virginia class ships, the Navy negotiated detail design separately from
construction,11 benefiting from the knowledge gained from detail design in
negotiating prices for construction. However, 2 years after negotiating
the detail design contract, the Navy negotiated target prices for the SSN
774 and
10 LPD 18 and LPD 19 were included in the contract as options to buy.
11 The practice followed with the San Antonio class ships of negotiating
detail design and lead ship construction together is a common Navy
practice. For example, over the next 3 years, the Navy's acquisition plans
call for awarding contracts covering both detail design and lead ship
construction for three new ship classes: DD(X) surface combatant, CVN 21
aircraft carrier, and the Littoral Combat Ship.
SSN 775, both considered lead ships for the two shipyards involved in
constructing submarines. Target prices for the first two follow-on ships,
SSN 776 and SSN 777 were agreed on at this time as well. As a result,
target prices for these follow-on ships did not benefit from the knowledge
gained from constructing the lead ships.
The practice of setting target prices early on affects not only the
realism of the contract target prices, but also the realism of the budgets
approved by the Congress to fund these contracts. In order to fund a
contract covering both detail design and lead ship construction,
authorization and funding for detail design and lead ship construction is
approved by the Congress in one budget year, before detail design begins.
For example, the Congress funded detail design and construction of LPD 17
in the fiscal year 1996 budget. While the follow-on ships, LPDs 19 and 20,
were funded in later years, budgets were still unrealistic because the
target prices were used as a basis for the budget request.
The size of the budget and the contract conditions can also affect the
realism of target prices. In negotiating the contract for the first four
Virginia class ships, program officials stated that the target price they
could negotiate was limited to the amount included in approved or planned
budgets. The shipbuilders said that they accepted a challenge to design
and construct these ships for $748 million less than their estimated costs
because the contract protected their financial risk. The contract included
a large minimum fee (profit), in addition to the incentive fee that would
be reduced in the event of cost growth. Moreover, the contract was
structured so that the Navy would pay the full cost of increases for
specialized, highly engineered components rather than share the cost
increases with the shipbuilder. The Navy also was responsible for the full
amount of growth in certain labor costs.
Recently, the Navy has supported the preparation of more realistic budget
requests. Program managers are encouraged to budget to their own estimate
of expected costs rather than at target prices that are not considered
realistic. For example, for the LPD 17, an acquisition decision memorandum
stated that the program will be budgeted to the Cost Analysis Improvement
Group estimate.12 Also, in negotiating recent
12 The Cost Analysis Improvement Group provides independent cost and risk
assessments and analyses of Major Defense Acquisition Programs for the
Office of the Secretary of Defense.
contracts for additional Virginia class and San Antonio class ships, the
Navy structured the contracts to encourage more realistic target prices.
Other Factors Affect Budget Realism
Beyond target prices, shifting priorities, and inflation accounting can
have a significant impact on the realism of ship budgets. Specifically,
budget requests are susceptible to across-the-board reductions to account
for other priorities, such as national security and changes in program
assumptions. Competing priorities create additional management challenges
for programs that receive a reduced budget without an accompanying
reduction in scope. For example, during the budget review cycles of 1996
through 2003, the initial cost estimate for DDGs 89-92 was decreased by
$119 million-or 55 percent of the total cost growth for the four DDGs. Had
the initial estimate not been reduced, the cost growth would have only
amounted to $96 million.
Inflation rates can also have a significant impact on ship budgets. Until
recently, Navy programs used Office of the Secretary of Defense and Office
of Management and Budget inflation rates.13 Inflation rates experienced by
the shipbuilding industry have historically been higher. As a result,
contracts were signed and executed using industry specific inflation rates
while budgets were based on the lower inflation rates, creating a risk of
cost growth from the outset. For the case study ships, the difference in
inflation rates, while holding all other factors constant, explains 30
percent of the $2.1 billion in cost growth for these ships.
In February 2004, the Navy changed its inflation policy directing program
offices to budget with what the Navy believes are more realistic inflation
indices. The Navy anticipates this policy change should help curtail
future requests for prior year completion funds.
13 The Office of the Secretary of Defense and Office of Management and
Budget inflation indices are based on a forecast of the implicit price
deflator for the Gross Domestic Product prepared by the Office of
Management and Budget and the White House Council of Economic Advisors.
The Gross Domestic Product includes all U.S. goods and services and is a
general economic indicator overarching many different commodities.
Cost Reporting Weaknesses Delayed Efforts to Mitigate Cost Risks
While DOD guidance allows some flexibility in program oversight, we found
that reporting on contractor performance was inadequate to alert the Navy
to potential cost growth for the eight case study ships. With the
significant risk of cost growth in shipbuilding programs, it is important
that program managers receive timely and complete cost performance reports
from the contractors. However, earned value management-a tool that
provides both program managers and the contractor insight into technical,
cost, and schedule progress on their contracts-was not used effectively.14
Cost variance analysis sections of the reports were not useful in some
cases because they only described problems at a high level and did not
address root causes or what the contractor plans were to mitigate them.
Earned value management provides an objective means to measure program
schedule and costs incurred. Among other requirements, DOD guidance on
earned value management requires that "at least on a monthly basis"
schedule and cost variances be generated at levels necessary for
management control. Naval Air Systems Command, which is considered a
center of excellence for earned value management, recommends that cost
performance reports be submitted at a minimum on a monthly basis, in part
to help the program manager mitigate risk. Officials from the command
stressed that because earned value management acts as an early warning
system, the longer the time lapse in receiving the cost performance
report, the less valuable the data become.
However, shipbuilders for the Nimitz and Virginia class ships we reviewed
submitted their official earned value management cost performance reports
to the Navy on a quarterly basis instead of monthly,15 delaying the
reports-and corrective action-by 3 to 4 months. Had the reporting been
monthly, negative trends in labor and materials on the Virginia class
submarine would have been revealed sooner and enabled corrective action to
occur quickly in areas of work that were not getting completed as planned.
Earlier reporting would have also alerted managers of cost performance
problems on the CVN 76 carrier. Because data on actual cost expenditures
for CVN 76 were provided incrementally and late, the program manager did
not identify a funding shortage until it was too late
14 For more information on the importance of Earned Value Management see
Appendix IV of our report GAO-03-600 entitled "Missile Defense: Additional
Knowledge Needed in Developing System for Intercepting Long-Range
Missiles."
15 Beginning in March 2006 the Navy will require monthly cost reporting
for CVN 77.
to remedy the problem. As a result, a contractwide stop-work order was
given. LPD 17 also experienced cost and schedule problems. To allow for
better tracking of schedule and costs and more timely response to
problems, the program manager changed the cost performance reporting
requirement from quarterly to monthly.
The quality of the cost performance reports, whether submitted monthly or
quarterly, was inadequate in some cases-especially with regard to the
variance analysis section, which describes any cost and schedule variances
and the reasons for these variances and serves as an official, written
record of the problems and actions taken by the shipbuilder to address
them. Both the Virginia class submarine and the Nimitz class aircraft
carrier programs' variance analysis reports discussed the root causes for
any cost growth and schedule slippage and described how these variances
were affecting the shipbuilders' projected final costs. However, the
remaining case study ship programs generally tended to report only
high-level reasons for cost and schedule variances with little to no
detail regarding root cause analysis or mitigation efforts16-making it
difficult for managers to identify risk and take corrective action.
Finally, the periodic reassessment of the remaining funding requirements
on a program and a good faith estimate at completion-another part of
earned value management-were inadequate to forecast the amount of
anticipated cost growth. Managers are required to evaluate the estimate at
completion and report it in the cost performance report, updating when
required. The Defense Contract Audit Agency recently observed the
importance of the shipbuilders' developing credible estimates at
completion and ensuring all estimates at completion revisions are
justified and made in a timely manner. However, the shipbuilders'
estimates for the study ships tended to be optimistic-that is, they fell
at the low end of our estimated cost growth range. Specifically,
shipbuilder estimates for four ships that are still under construction
were near our low estimate, (See fig. 3), leading management to believe
that the ships will cost less than what is likely to be the case.
16 A recent Defense Contract Audit Agency audit found that while one
shipbuilder identified material cost and schedule variances in its
variance analysis report, it did not provide written documentation related
to the reasons for the variance or provide explanation for the variances
in the cost performance report.
Figure 3: GAO and Shipbuilder Construction Cost Growth Forecasts for Case
Study Ships
Ship
CVN 77
(Nimitz class aircraft carrier)
LPD 17
(San Antonio class amphibious transport dock ship)
LPD 18
(San Antonio class amphibious transport dock ship)
SSN 775
(Virginia class submarine)
0
In millions of dollars
Conclusions
200 400 600 800 1,000
Navy's FY 2005 budget
Range of GAO's forecast for additional cost growth for construction
Shipbuilder's estimate for additional cost growth for construction
Source: Shipbuilder (data); GAO (analysis).
See appendix VI for more details on the cost growth forecasts for ships
currently under construction.
The challenge in accurately estimating and adequately funding the
construction of Navy ships is framed by the long construction time cost
estimates must account for and the fact that ships must be fully funded in
the first year of their construction. Thus, an underestimation of costs, a
budget reduction, or an increase in cost, creates a need for additional
money that must be requested and appropriated. The fact that requests have
been sizable and have occurred routinely over the years suggests that the
Navy can do better in getting a match between the estimated costs of new
ship construction and the money it budgets to pay for them. The goal is
not necessarily to eliminate all requests for additional funds, for that
could lead to overbudgeting or deferring necessary design changes. Rather,
the goal is to get a better match between budgeted funds and costs in
order that the true impact of investment decisions is known.
Our work shows that currently, the Navy's cost estimating, budgeting, and
contracting practices do not do a good enough job of providing for the
likely costs of building ships. This is particularly true for first of
class
ships, for which uncertainty is highest. Moreover, when actual costs begin
to go astray of budgeted funds, management tools intended to flag
variances and enable managers to act early are not always effectively
employed. If these practices are to lead to more realistic results---and
reduced overruns-they will have to produce and take advantage of higher
levels of knowledge. In some cases, improved techniques, such as
performing uncertainty analyses on cost estimates, can raise the level of
knowledge. In other cases, such as contracting for detail design and
construction on first-of-class ships, contracting in smaller steps can
allow necessary knowledge to build before major commitments are made.
The Navy has recognized the need to get a better match between funding and
cost and is providing guidance to achieve this match. The success of this
guidance will depend on how well it is implemented on individual programs.
There are additional steps the Navy can take, which are detailed in our
recommendations. Taking these steps now is especially important for the
Navy as it embarks on a number of new, sophisticated shipbuilding
programs. If a better match between funding and cost is not made,
additional funds needed for cost growth will continue to compete for the
funds needed for new investments in ships or other capabilities. Difficult
budget choices are ahead making it essential that priorities are set with
a clear understanding of the financial implications of different spending
and investment alternatives. To the extent unplanned demands on the budget
can be reduced, better informed decisions can be made.
We are recommending that the Secretary of Defense take the following seven
actions.
To improve the quality of cost estimates for shipbuilding programs and
reduce the magnitude of unbudgeted cost growth, we recommend the Secretary
of Defense
o conduct independent cost reviews for all follow-on ships when
significant changes occur in a program and establish criteria as to what
constitutes significant changes to a shipbuilding program,
o conduct independent reviews of every acquisition of an aircraft
carrier, and
o direct the Secretary of the Navy to develop a confidence level for all
ship cost estimates, based on risk and uncertainty analyses.
Recommendations
To assure that realistic prices for ship construction contracts are
achieved, we recommend that the Secretary of Defense direct the Secretary
of the Navy to
o negotiate prices for construction of the lead ship separately from the
pricing of detail design and
o separate pricing of follow-on ships from pricing of lead ships,
negotiating prices for early ships in the budget year in which the ship is
authorized and funded.
To improve management of shipbuilding programs and promote early
recognition of cost issues, we recommend that the Secretary of Defense
direct the Secretary of the Navy to
o require shipbuilders to submit monthly cost performance reports and
o require shipbuilders to prepare variance analysis reports that identify
root
Agency Comments
and Our Evaluation
causes of reported variances, associated mitigation efforts, and future
cost impacts.
DOD agreed with our recommendations to conduct independent reviews of
every aircraft carrier and to develop a confidence level for all ship cost
estimates, based on risk and uncertainty analysis. DOD partially agreed
with our recommendations about contract pricing and cost performance
reporting-areas the Navy noted it has taken some measures to improve.
While the Navy has taken steps in the right direction, we believe more
must be done to reduce ship cost overruns, consistent with our
recommendations.
We made a recommendation in our draft report that independent reviews be
conducted for all follow-on ships when significant changes to the program
occur. DOD responded that it will request additional assessments, if
needed after Milestone B. It is important that criteria be established for
determining when additional assessments are needed. Programs may undergo
several changes after the required estimate, such as the Arleigh Burke
destroyer, which underwent four major upgrades since its only independent
estimate in 1993. We believe DOD needs to establish criteria concerning
what significant changes to a program trigger an independent cost estimate
and have modified our recommendations accordingly. DOD could clarify
whether these changes include baseline, profile, or major systems
upgrades, for instance.
DOD stated that it will consider, on a case-by-case basis, negotiating
detail design separately from the lead ship and negotiating early
follow-on ships separately from the lead ship. We believe that this
approach should be the normal policy, if overruns are to be reduced. Ships
represent a substantial investment--more than $1 billion for each
destroyer and amphibious transport, about $2.5 billion for the lead ship
in the next class of destroyers, $2.5 billion for submarines, and several
billion for carriers. Ships costing substantially less-for example, $220
million for each Littoral Combat Ship-are the exception rather than the
norm. A realistic target price is important for structuring contract
incentives and providing informed budgets to the Congress. Deciding prices
for the lead ship and follow-on ships together before detail design has
even begun on the lead ship is unlikely to yield realistic prices. Insight
gained into material costs and labor effort even in the first year of
detail design will make realistic pricing of the lead ship more feasible.
Similarly, experience gained in the first years of construction can
benefit the realism of prices for follow-on ships.
DOD noted that the Navy is already requiring shipbuilders to submit cost
performance reports monthly with one exception. With the Nimitz class
program beginning monthly reporting in March 2006, the Virginia class will
be the only program to submit quarterly instead of monthly cost
performance reports. DOD states that the Navy has access to labor hour
data in the interim. While informal access to timely data is preferable to
delayed access, without written, formal cost reporting there is less
visibility or accountability from the last formal report to the next cost
performance report 3 months later. The Virginia class program has
experienced significant cost increases and experienced one of the largest
prior year funding requests of programs we reviewed. LPD 17 and carrier
program officials recognized that more frequent formal reporting and
review of cost performance helped them to better manage cost growth and
changed their program reporting requirements from quarterly to monthly.
Although variance analysis reporting is required as part of cost
performance reporting and is being conducted by the shipbuilders, we
observed that there is wide variation in the quality of these reports. DOD
rightly observes that these reports are one of many tools used by the
shipbuilders and DOD to track performance. To be a useful tool, however,
we believe it is important that shipbuilders provide the government with
detailed analyses of the root causes and impacts of cost and schedule
variances. Cost performance reports that consistently provide thorough
analysis of the causes of variances, their associated cost impacts, and
mitigation efforts will allow the Navy to more effectively manage, and
ultimately reduce, cost growth.
DOD's detailed comments are provided in appendix VII.
As agreed with your office, unless you announce its contents, we will not
distribute this report further until 30 days from its date. At that time,
we
will send copies to the Secretary of Defense, the Secretary of the Navy
and
interested congressional committees. We will also make copies available
to others upon request. In addition, the report will be available at no
charge on the GAO Web site at http://www.gao.gov.
If you or your staff have any questions concerning this report, please
contact me at (202) 512-4841 or Karen Zuckerstein, Assistant Director, at
(202) 512-6785. Key contributors to the report are identified in
appendix VIII.
Sincerely yours,
Paul L. Francis
Director
Acquisition and Sourcing Management
Appendix I: Scope and Methodology
Our methodology for all three objectives included a case study analysis of
eight ships. These ships were in four ship classes: Virginia class
submarines, LPD 17 amphibious assault ships, Arleigh Burke destroyers, and
Nimitz class carriers. We selected these ship classes and these ships
based on data contained in the "Naval Sea Systems Command Quarterly
Progress Report for Shipbuilding and Conversion Status of Shipbuilding
Programs," dated July 1, 2003. This report identifies all ships under
construction and the progress in terms of "percent complete" for each
ship. We looked only at new construction and excluded ship conversions.
This report identified eight ship classes with ships under construction.
In addition to the four ship classes that we studied, the report
identified ships in the Seawolf attack submarine, LHD amphibious assault
ship, T-AKE cargo ship, and T-AKR vehicle cargo ship classes. We did not
review the Seawolf and T-AKR ship classes because construction of these
classes was at an end and were unlikely to affect future budgets. We did
not include ships from the remaining two classes because we limited the
ship selection to those ships that were more that 30 percent complete and
none of the ships in those two classes met those criteria.
We selected two ships per class for the four classes we reviewed. Where
possible, we chose a lead and follow-on ship. We also looked at which
shipyards were building these ships in order to get coverage of the major
shipyards. We limited the selection to ships more than 30 percent complete
so we had sufficient information on program performance. Three
Virginia-class submarines, three amphibious assault ships, two carriers,
and 12 destroyers met this criterion. For the Virginia class program, we
initially chose SSN 774 and SSN 776; both built and integrated at the
Electric Boat, Connecticut shipyard.1 As we gained knowledge of the
program and Newport News' role in constructing and launching half of the
submarines in this class, we substituted the SSN 775 for the SSN 776.
Characteristics of the ships we selected are summarized in table 12.
1 Under a teaming arrangement, two-thirds of the SSN 774 and 776 is
constructed at Electric Boat with the remaining third built at Newport
News and shipped to Electric Boat for final assembly. For the SSN 775, the
inverse is true.
Table 12: Characteristics of Case Study Ships
Amphibious Virginia class Nimitz class Arleigh
Burke
assault ships submarines carriers destroyers
LPD 17 LPD 18 SSN 774 SSN 775 CVN 76 CVN 77 DDG 91 DDG
92
Avondale Electric Newport News Ingalls Bath
Shipyard Avondale Newport News Newport News Iron
Operations Boat Shipbuilding Operations
Operations Shipbuilding Shipbuilding Works
Percent 86 43 87 79 99 35 96
complete as of
July 2003
Lead Follow Lead Leada Follow Follow Follow
Lead/follow Follow
Source: Navy (data), GAO (presentation)
aSSN 775 is considered the lead ship for Newport News.
Because a large percentage of the ship construction budget is allocated to
fund the shipbuilding contracts, we assessed the shipbuilders' cost
performance for the four classes of ships in our study. To make these
assessments, we applied earned value analysis techniques to data captured
in shipbuilder cost performance reports. We also developed a forecast of
future cost growth. For ships currently under construction (and more than
30 percent complete), we compared the initial target costs with the likely
costs at the completion of the contracts using established earned value
formulas. We based the lower end of our cost forecast range on the costs
spent to date added to the forecast cost of work remaining. The remaining
work was forecast using the cumulative cost performance index efficiency
factor. Studies have shown that using this method is a reasonable estimate
of the lower bound of the final cost. For the upper end of our cost range,
we relied on either the actual costs spent to date added to the forecast
of remaining work with an average monthly cost and schedule performance
index or a cost/percent complete trend analysis, whichever was higher.
In order to understand the components of cost growth, we used cost data
provided by the shipbuilders for each of the case study ships. In most
cases we compared the initial target cost to the current target cost. As a
result some of the increases in target cost could have resulted from
additional contract modifications initiated by the Navy, cost overruns due
to the shipbuilder, or unanticipated events. Most shipbuilders allocate
contract costs into three categories: material costs, labor costs, and
overhead costs. We, however, used these data to allocate costs into the
following categories: labor hours, material costs, and labor and overhead
rates. Since labor costs and overhead costs can change due to labor hours,
and labor and overhead rates, we separated the program overhead cost
associated with an increase in labor and overhead rates from the program
overhead cost associated with an increase in labor hours. This was
accomplished by holding each component constant to isolate the impact.
After we isolated the program overhead cost associated only with
additional labor hours, we added this to the shipbuilders' reported labor
cost growth and subtracted this from the shipbuilders' reported overhead
cost growth. Our analysis captures all costs associated only with overhead
and labor rate changes. Increases in overhead related to growth in labor
hours are captured only in our analysis of labor hour increases.
We used the latest cost performance data available to us in July 2004. The
latest available cost performance reports for the case study ships were as
follows
o DDG 91 June 2004,
o DDG 92 May 2004,
o CVN 76 July 2003,
o CVN 77 March 2004,
o LPD 17 and LPD 18 May 2004, and
o SSN 774 and SSN 775 July 2004.
In order to understand the funding and management practices that
contribute to cost growth, we reviewed Navy acquisition guidance and
reviewed best practices literature for weapons systems construction. To
better understand the budgeting of ships and the acquisition process we
met with officials at the Navy and Office of Secretary of Defense
Comptroller. Based on indicators from our case study analysis that cost
estimating practices may contribute to cost growth, we met with cost
estimators, including those from Naval Sea Systems Command, Cost Analysis
Improvement Group, and the Navy Cost Analysis Division. We reviewed DOD
and Navy cost estimating policies, procedures, and guidance. Additionally,
we met with cost estimators from the Naval Air Systems Command, the Air
Force, and the Army to compare how Naval Sea Systems Command estimating
practices vary from other military cost estimating practices. We
interviewed program officials, contracting officers, and shipbuilders and
reviewed shipbuilder reports, which included explanations for cost growth.
We met with officials at Supervisor of Ships and the Defense Contract
Audit Agency both at the shipyards and at headquarters to review their
oversight policies, procedures, and practices. We met with Navy Audit
Service officials to gain information on earned value management reviews
at shipyards. We also reviewed contract documentation and audit reports.
Our analysis relied on shipbuilders' earned value data. To establish the
reliability of the data, we examined the integrated baseline reviews that
are conducted at the beginning of a contract. We also confirmed that the
shipbuilders had validated earned value systems.
We performed our review from July 2003 to December 2004 in accordance with
generally accepted government auditing standards.
Appendix II: Arleigh Burke Class Destroyer
Program Description The USS Arleigh Burke class destroyer (DDG 51)
provides multimission offensive and defensive capabilities, and can
operate independently or as part of carrier strike groups, surface action
groups, and expeditionary strike groups. The DDG 51 class, which is
intended to replace earlier surface combatant classes, was the first U.S.
Navy ship designed to reduce radar cross-section and its detectability and
likelihood of being targeted by enemy sensors and weapons. Originally
designed to defend against Soviet aircraft, cruise missiles, and nuclear
attack submarines, the ship is to be used in high-threat areas to conduct
antiair, antisubmarine, antisurface, and strike operations.
As of May 2004, 43 Arleigh Burke destroyers have been delivered to the
Navy, with a total of 62 to be delivered at the end of the production.
Funding for the lead ship (DDG 51) was provided in fiscal year 1985. The
lead ship construction contract was awarded to Bath Iron Works in April
1985. With the award of the follow-on ship-DDG 52-to Ingalls Shipbuilding
Incorporated-a second shipbuilder was established. The DDG 91 and DDG 92,
which are covered in this report, include a number of upgrades, such as
the most current Aegis weapons system; installation of a remote
mine-hunting system capability and the introduction of commercially built
switchboards.
Figure 4: Arleigh Burke Class Destroyer
Source: Navy.
Table 13: Major Events in the Acquisition of DDG 91 and DDG 92
May 1996 o Initial estimate for three ships developed using actual
costs from DDG 70-the last ship of the prior configuration in preparation
for the 1998 President's budget.
September 1996 o Congress authorizes procurement of 4 DDGs-DDGs 89-92.
August 1997 o Competition to determine contract awards. Navy uses
"Profit Related to Offers" strategy, where the lower offeror receives a
larger proportion of the contract's target profits, while the higher
offeror receives a smaller proportion. Both shipbuilders stated that they
proposed aggressively low costs in an effort to win the higher profit
margin on the contract that included DDG 91 and DDG 92.
January 1998 o Fiscal Year 1999 Presidential Budget submission is $119
million less than the initial budget estimate due to across the board DOD
and Office of Management and Budget reductions.
March 1998 o Bath Iron Works receives a contract for the DDG 92 with an
initial target price of $351 million. Ingalls Shipbuilding receives a
contract for the DDG 91 with an initial target price of $355 million.
September 2000 o DDG 91 construction begins.
December 2000 o DDG 92 construction begins.
April 2001 o Northrop Grumman Ship Systems acquired Ingalls
Shipbuilding.
June 2002 o Northrop Grumman Ship Systems and Bath Iron Works agree to
a swap in which LPDs (including LPD 19) will be built by Northrop Grumman
and future DDG ships will be built at Bath Iron Works.
October 2003 o DDG 91 delivered.
May 2004 o DDG 92 delivered.
Sources: Navy and shipbuilders (data).
Cost Experience on DDG 91 and DDG 92 cost $135 million more than budgeted.
(See table 14.) The Congress has appropriated almost $100 million to cover
these
DDG 91 and DDG 92 increases.
Table 14: Growth in Program Budgets for Case Study Ships
Dollars in millions
Initial and FY 2005 President's budget Difference in budgets Case study ship
Initial FY2005 Total difference
Difference due to Navy-furnished equipment Difference due to construction
costsa
DDG 91 $917 $997 $80 $43 $37
DDG 92 925 979 55 (7)b
Total $1,842 $1,976 $135 $36 $99
Main Drivers of Cost Growth for DDG 91 DDG 92
Sources: Navy (data); GAO (presentation).
aPart of the increased cost is due to changes in the scope of the
contract.
bNegative reflects savings resulting from the use of a more economical
warfare system than was initially budgeted on the DDG 92.
Construction costs-especially the costs associated with the number of
labor hours needed to build the ships-were the major source of cost
growth. Navy-furnished equipment, including the Aegis weapon system, was
also a significant source of cost growth for the two DDGs, representing 21
percent of the cost growth. Increases in the number of labor hours account
for 67 percent of the cost growth on shipbuilding construction contracts.
We found that ship overhead-such as employee benefits and shipyard support
costs-and labor rate increases accounted for 21 percent of cost growth.
The two DDGs actually underran material costs, due to DDG 91 material cost
savings.
Figure 5: Average Sources of Cost Growth on DDG 91 and DDG 92
Labor hours
Material -9
Overhead and labor rates
Navy-furnished equipment
-100 102030405060
Percent of cost growth
Percentage of overall cost growth due to shipbuilder construction costs
Percentage of overall cost growth due to cost of Navy-furnished equipment
Labor Hours
Labor hour increases account for the majority of the cost growth on DDG 91
and DDG 92. (See fig. 5.) DDG 91 required almost 1 million hours of
additional labor hours and DDG 92 required an additional 2 million hours
above the original contract proposal. DDG 91 and DDG 92 incorporated a
number of new technologies in their design, including the remote
mine-hunting system, which consists of a remote operated vehicle and a
launch and recovery system stored within the ship. To accommodate this
system, designers had to make significant structural changes to 26 of the
ship's 90 design zones. When construction began on DDG 91 and DDG 92, the
remote mine-hunting system's design was not mature. As a result,
significant details of the design could not be captured in the
shipbuilders' planned contract costs. Moreover, the shipbuilders
anticipated that the system's design would be completed in July 1999-
several months before the start of ship fabrication in November 1999.
However, it was not completed until November 2001, with additional
revisions to the design occurring through March 2003. Because the design
was changing as installation of the system began, laborers re-installed
parts of the system, increasing the engineering and production hours.
As the number of hours to construct the ship increased, total labor costs
grew, with the shipbuilder paying for additional employee wages and
overhead costs. As table 15 shows, we separated the overhead and labor
rates associated with the additional hours and added this to the
Source: Navy (data); GAO (presentation).
shipbuilders reported labor cost growth. Our analysis thus captures all
costs associated with labor hour growth-including overhead and labor
rates. The methodology we used to separate the overhead costs associated
with rate increases and labor hour increases is discussed in appendix 1.
Table 15: Growth in Labor Hour Costs
Dollars in millions
Analysis based on June 2004 (DDG 91) and May 2004 (DDG 92) data
Overhead and Labor hour
labor cost as a
Shipbuilder rate dollars on Total dollars percent of
reported due to total
Case study labor dollar increased labor increased labor contract
ship growth hours hours growth
DDG 91 $23 $24 $47 105%
DDG 92 43 42 85
Total $66 $66 $132
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events.
According to the shipbuilder, additional labor hours were also needed to
complete DDG 91 because many experienced workers had left the trade in
favor of higher paying jobs in the area and, as a result, less experienced
workers took longer to finish tasks and made mistakes that required
rework. For DDG 92, workers encountered challenges in building the ship
due to a new transfer facility that enabled the shipyard to construct a
greater proportion of the ship on land. The ship was constructed using
larger subsections or units. While the shipbuilder expects that the
facility will improve efficiency, on DDG 92, workers had to learn new
processes and had difficulties aligning larger units of the ship to one
another. Labor hours increased as workers spent additional time realigning
and combining the units to make larger sections of the ship.
Navy-Furnished About $38 million of Navy-furnished equipment cost growth
is associated
Equipment with the Aegis weapon system, specifically the purchase of an
additional SPY-D radar used in system testing.1 The Navy originally
planned to move the developmental radar from the engineering and
development site to the final testing and certification center. However,
increased complexity
1 These costs are non-recurring, affecting only DDG 91 costs.
involved with the introduction of a new radar and new computing plant
required more development time than was originally planned. In order to
ensure timely delivery of DDG 91, the Navy procured a second radar for the
testing facility, allowing the Navy to simultaneously finish final
development of the radar, while at the same time beginning testing and
certification of the Aegis weapon system computer program.
Overhead Rate Increases Our analysis shows that program overhead costs
and increases in labor rates accounted for approximately 21 percent of the
cost growth on the DDG 91 and 92 contracts. Table 15 includes overhead
increases that were a consequence of labor hour increases. Table 16
isolates the remaining portion of overhead increases due to increases in
rates.
Table 16: Growth in Overhead Costs and Labor Rates
Dollars in millions
Analysis based on June 2004 (DDG 91) and May 2004 (DDG 92) data
Overhead cost from rate
Shipbuilder reported Portion related to growth Portion related to
increases as a percent of
Case study ship overhead growth in labor hours overhead and labor rates
total contract growth
DDG 91 $43 $24 $20 44%
DDG 92 56 42 14
Total $99 $66 $34
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events.
Despite savings incurred through the consolidation of Ingalls Shipyard
into Northrop Grumman,2 overhead rates were about 13 percent higher than
anticipated in 2001. According to shipbuilders, increases in overhead
rates can be attributed largely to changes in the shipyard's workload and
employee benefit costs. After the cancellation of the construction
contract for a commercial cruise ship due to the company's bankruptcy and
the delay in signing the contract for the next generation destroyer,
overhead costs had to be absorbed by the remaining contracts at the yard,
including the DDG 91. Similarly, on DDG 92, the shipbuilder based its
overhead rates on anticipated work from the construction of the next
generation
2 According to the shipbuilder, consolidation of Ingalls shipbuilding and
Avondale shipyards into Northrop Grumman reduced overhead costs by
approximately 3 percent.
destroyer and the San Antonio class ships. When these programs did not
materialize as expected, the other programs at the yard assumed overhead
costs. At both shipyards health and dental care costs increased. For
example, at one shipyard, the shipbuilder negotiated a favorable medical
insurance contract but the insurance company went bankrupt, forcing the
shipbuilder to become self-insured-at a higher cost.
Both shipbuilders were also affected by labor rate increases. Following a
strike at Bath Iron Works, the union negotiated a $1.12 increase in labor
rates or a $6 million increase above the costs projected in the contract.
For Northrop Grumman, between the initial proposal and the latest
estimate, the labor rate increased by $1.50 per hour for a total impact on
the DDG 91 of $7 million.
Material Costs As shown in table 17, material cost increases did not
represent a major source of cost increases for DDG 91 and DDG 92-largely
because the materials were purchased for four ships at one time.
Table 17: Growth in Material Costs
Dollars in millions
Analysis based on June 2004 (DDG 91) and May 2004 (DDG 92) data
Material cost as a percent
of
Case study Increased material Percent increase total contract growth
ship costs
DDG 91 ($22) (13%) (49%)
DDG 92 30 20
Total $8
Sources: Navy (data); GAO (analysis).
However, DDG 92 overran its material budget by $30 million-73 percent of
which was due to information technology, small tooling, and other material
costs. Although these costs comprise only 17 percent of the material cost
budget, their costs are driven by labor hour usage-as additional labor
hours were needed to construct DDG 92, additional tools were needed,
raising material costs. Material costs also increased because the
shipbuilder began allocating information technology costs to materials-not
overhead, as it had initially done. DDG 91 experienced a $22 million
underrun of material costs. According to the shipbuilder, the underrun was
due to efficiencies gained through the consolidation of Ingalls Shipyard
with nearby Avondale Shipyard-also owned by Northrop Grumman Ship Systems.
With the consolidation, Northrop Grumman
stated it could purchase materials for both shipyards-creating cost
savings that were not anticipated in DDG 91's original material cost
budget.
Appendix III: Nimitz Class Aircraft Carrier
Program Description The mission of the Nimitz class nuclear powered
aircraft carriers-which are intended to replace the Navy's conventionally
powered carriers-is to provide a sustained presence and conventional
deterrence in peacetime; act as the cornerstone of joint allied maritime
expeditionary forces in crises; and support aircraft attacks on enemies,
protect friendly forces, and engage in sustained independent operations in
war. Nine Nimitz class nuclear carriers-CVN 68 through CVN 76-have been
delivered since acquisition of the first ship in October 1967. CVN 77, the
tenth and final ship of the class, is a modified version of CVN 76 and
will serve as a transition ship to the next generation of aircraft
carriers. Both CVN 76 and CVN 77 included several significant design
changes, including a bulbous bow; larger air-conditioning plants; a
redesigned island; weapons elevator modifications; and an integrated
communications network.
Figure 6: Nimitz Class Aircraft Carrier
Source: Navy.
Table 18: Major Events in the Acquisition of CVN 76 and CVN 77
August 1993 o Initial estimate for CVN 76 developed based on CVN 74/75
proposal data with adjustments made for design and economic factors.
September 1994 o Congress appropriates funding for the construction of
CVN 76.
December 1994 o Navy awards a fixed-price incentive fee contract for
detailed design and construction of CVN 76 to Newport News Shipbuilding
for a target price of $2.5 billion.
May 1995 o Construction of CVN 76 begins.
April 1999 o 4-month strike at Newport News causes work stoppage on
construction of CVN 76.
July 2000 o Congress appropriates funding for the construction of CVN
77.
January 2001 o Navy awards a fixed-price incentive fee contract for
detailed design and construction of CVN 77 to Newport News Shipbuilding
for a target price of $3.4 billion.
February 2001 o Navy requests $86 million for CVN 76 and $20 million
for CVN 77 in prior year funding to cover cost growth.
June 2001 o Construction begins on CVN 77. Advanced construction began
in 1998 to provide sustaining work for the shipyard.
February 2002 o Prior year request of $94 million for CVN 76 and $75.4
million for CVN 77.
September 2002 o Development delays prompt Navy to revert to a legacy
warfare system on CVN 77. Costs for warfare system are transferred from
the shipbuilder to the Navy.
December 2002 o Original contract delivery date for CVN 76.
June 2003 o CVN 76 delivered to the Navy-6 months later than the
original delivery date.
August 2003 o Navy initiates a contract wide stop work order on CVN 76
to prevent depletion of program funding. Stop work order is rescinded 3
months later.
February 2005 o Navy requests $870 million in prior year funding over
fiscal years 2006 to 2008 for CVN 77. This includes an increase in $908
million in construction costs and a $38 million decrease in Navy-furnished
equipment.
March 2008 o Initial expected delivery date for CVN 77.
January 2009 o Current expected delivery date for CVN 77.
Cost Experience on
CVN 76 and CVN 77
Sources: Shipbuilder and Navy (data), GAO (presentation).
The Fiscal Year 2005 President's Budget showed that budgets for the CVN
case study ships had increased by $173 million, and the Congress has
appropriated funds to cover these increases. However, based on March 2004
data, we projected additional cost growth on contracts for the carriers is
likely to reach $485 million and could be higher. Therefore, the Navy will
need additional appropriations to cover this cost growth.
The fiscal year 2005 budget for the carriers is about $9.6 billion- $173
million more than the initial budget request for these ships. (See table
19.) As a result, the Navy has requested $275.4 million through both the
prior year completion bill and other financial transfers to fund cost
increases on the CVN program.1 Ship construction costs comprise the
majority of this increase.
On CVN 76, ship construction costs grew by $252 million above the initial
budget. As a result of cost growth, CVN 76 was in danger of running out of
funding. The program office issued over 75 stop work orders-including one
contract wide stop work order to temporarily save funding. Lower priority
work was cancelled or halted to avoid further cost growth. While stop work
orders saved money in the short term, they resulted in significant costs
later. On CVN 76 some work had to be completed in a post-delivery
contract-at a higher cost.
Table 19: Growth in Program Budgets for Case Study Ships
Dollars in millions
Initial and FY 2005 President's budget Difference in budgets Case study ship
Initial FY 2005 Total difference
Difference due to Navy-furnished equipment
Difference due to construction costsa
CVN 76 $4,476 $4,600 $124 ($128) $252
CVN 77 4,975 5,024 49 100 (51)
Total $9,451 $9,624 $173 $28 $201
Sources: Navy (data); GAO (presentation).
aPart of the increased cost is due to changes in the scope of the
contract.
We calculated a range of the potential growth for CVN 77 and found that
the total projected cost growth is likely to exceed $485 million and could
reach $637 million.2 (See table 20.)
1 $180 million was for CVN 76 and $95.4 million for CVN 77.
2 We did not project cost growth for CVN 76 because the ship has been
delivered to the Navy.
Table 20: GAO's Forecasts of Additional Cost Growth for Construction
Dollars in millions
Analysis based on July 2003 (CVN 76) and March 2004 (CVN 77) data
Amount already
requested to GAO's forecast
cover for
Percent of ship contractor's additional cost GAO's forecast
growth of total
Case study construction increased cost for construction cost growth
ship completed
CVN 76 Delivered $252 $0-0 $252-252
CVN 77 45% (51)a 485-637 434-586
Total $201 $485-637 $686-838
growth
Sources: Navy (data); GAO (analysis).
Note: Cost growth is from original contract target price, not from the
current contract target price. Forecast reflects expected price to the
Navy.
aNegative reflects shifting of funds from the construction contract to
Navy-furnished equipment.
Our cost growth estimates have proven to be understated. The Fiscal Year
2006 President's Budget recognizes cost growth of $908 million for ship
construction above the prior year's budget request. In addition, we
assumed that the shipbuilder will maintain its current efficiency through
the end of the contracts and meet scheduled milestones. For example, Navy
officials told us that delivery of CVN 77 is likely to slip to January
2009, further increasing the final cost of the ship.
Based on 2004 data, increases in labor hour and material costs account for
80 percent of the cost growth on CVN 76 and CVN 77, while the costs for
Navy-furnished equipment-including propulsion and weapon systems-
declined.3 (See fig. 7.) The remaining 23 percent of cost growth resulted
from increases in overhead costs. The shipbuilder cited a number of direct
causes for the labor hour, material, and overhead cost growth in the case
study ships. The most common causes were related to demands for labor on
other programs at the shipyard, the need for additional and more costly
materials, and changes in employee pay and benefits.
Main Drivers of Cost Growth for CVN 76 and CVN 77
3 Analysis is based on data available through March 2004.
Figure 7: Average Sources of Cost Growth on CVN 76 and CVN 77
Labor hours 40
Material 40
Overhead and labor rates
Navy-furnished equipment -3
0 102030405060 Percent of cost growth
Percentage of overall cost growth due to shipbuilder construction costs
Percentage of overall cost growth due to cost of Navy-furnished equipment
Source: Navy (data); GAO (presentation).
Materials Material costs increased on CVN 76 by $294 million and on CVN
77 by $134 million since the contracts were first awarded.
Table 21: Growth in Material Costs
Dollars in millions
Analysis based on July 2003 (CVN 76) and March 2004 (CVN 77) data
Total dollars due to Material cost as a
increased percent
Case study material costs Percent increase of total contract
ship growth
CVN 76 $294 43% 46%
CVN 77 134 13
Total $428
Sources: Navy (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events.
On both CVN 76 and CVN 77, material costs grew, in part, because the
shipbuilder underestimated the original budget for materials. In April
2002-7 years after construction began on CVN 76-about $32 million in
errors in material purchase estimates were revealed. CVN 77 has also
experienced a significant increase in material costs due to under
budgeting. According to the shipbuilder, a compressed construction
schedule on CVN 77 resulted in the budget for materials being established
prior to the completion of the carrier's design and even the completion of
design work on certain systems on CVN 76. As a result, the true magnitude
of the carrier's material costs was not known at the time of the contract
negotiation. Early in CVN 77 construction, however, the shipbuilder
reassessed the materials needed for construction in order to have a more
realistic estimate of final material costs. The Navy and the Defense
Contract Audit Agency recognized the absence of needed information on
materials during its review of the shipbuilder's proposal and expressed
concerns about the adequacy of the cost estimating system. According to
Newport News officials, the shipyard and Defense Contract Audit Agency are
working to resolve their concerns. The shipbuilder is estimating $200
million in material cost increases and additional funds are being
requested to cover this increase.
According to the shipbuilder, material cost increases on both CVN 76 and
CVN 77 can be attributed to increases resulting from a declining supplier
base and commodity price increases. Both carriers' material costs have
been affected by an over 15 percent increase in metals costs, which, in
turn, increases costs for associated components used in ship construction.
Moreover, many of the materials used in the construction of aircraft
carriers are highly specialized and unique-often produced by only one
manufacturer. With fewer manufacturers competing in the market, the
materials are highly susceptible to cost increases.
Other reasons for material cost increases include the following:
CVN 76
o Expenses of about $20 million in non-nuclear engineering effort that
were subcontracted for in late 1997 and of about $50 million for
information services were transferred from overhead to material in the
middle of the project.
CVN 77
o The expansion of commercial-off-the-shelf equipment in CVN 77 resulted
in additional costs to test the materials to make sure military
specifications were met.
Labor Hours
Costs on both carriers grew because of additional labor hours required to
construct the ships. At delivery, CVN 76 required 8 million hours of
additional labor hours to construct, while CVN 77 has required 4 million
hours. As the number of hours to construct the ship increased, total labor
costs grew, with the shipbuilder paying for additional employee wages and
overhead costs.
Table 22: Growth in Labor Hour Costs
Dollars in millions
Analysis based on July 2003 (CVN 76) and March 2004 (CVN 77) data
Overhead and Labor hour
labor cost as a
Shipbuilder rate dollars on Total dollars percent of
reported due to total
Case study labor dollar increased labor increased contract
ship growth hoursa labor hours growth
CVN 76 $78 $144 $222 35%
CVN 77 75 107 182
Total $153 $251 $404
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. We separated the overhead and labor rates associated
only with the additional hours and added this to the shipbuilders'
reported labor cost growth. Our analysis captures all costs associated
with labor hour growth-including overhead and labor rates.
aContractor performance reports included $63 million in overhead costs for
CVN 76 and $40 million for CVN 77 that have been disallowed (not charged
to the government).
Increases in labor hours were due in part to an underestimation of the
labor hours necessary to construct the carriers. The shipbuilder
negotiated CVN 76 for approximately 39 million labor hours-only 2.7
million more labor hours than the previous ship-CVN 75. However, CVN 75
was constructed more efficiently because it was the fourth ship of two
concurrent ship procurements. (See table 23.) CVN 76 and CVN 77, in
contrast, were procured as single ships. As table 23 shows, single ship
procurement is historically less efficient than two ship procurements,
requiring more labor hours.
Table 23: Historical Man-hours Used to Produce Prior Ships Compared to CVN 76
Negotiated Man-hours
Man-hours in millions
Hull Total man-hours Labor-hour change Type of ship Contract award
buy date
CVN 70 36.4 0 Single April 1974
CVN 71 44.3 7.9 Single September 1980
CVN 72 42.7 (1.6) Two December 1982
CVN 73 38.2 (4.5) Two December 1982
CVN 74 36.9 (1.3) Two July 1988
CVN 75 36.3 (0.6) Two July 1988
CVN 76 39.0 2.7 Single December 1994
Sources: Navy, Shipbuilder (data); GAO (analysis).
The shipbuilder and Navy budgeted the same number of hours to construct
CVN 77 as to construct CVN 76, despite forecasts showing that at 55
percent complete CVN 76 would need almost 2 million more hours above the
negotiated hours to complete the ship. To date, CVN 77 is expected to
incur over 4 million man-hours more than negotiated.
Some of the labor hour increase on CVN 76 occurred as a result of demands
for labor on other programs at the shipyard. During construction of CVN
76, 1 million hours of labor were shifted from the construction of the
carrier to work on the refueling and overhaul of CVN 68. The Navy deemed
the carrier overhaul and refueling effort as a higher priority than new
ship construction because carriers were needed back in the fleet to meet
warfighting requirements. Many of the most skilled laborers were moved to
the refueling effort, leaving fewer workers to construct CVN 76. Without
many of the necessary laborers to construct the ship, the CVN 76
construction schedule was delayed. In order to meet construction schedule
deadlines employees were tasked to work significant overtime hours.
Studies have shown, however, that workers perform less efficiently under
sustained high overtime.
Problems with late material delivery also led to labor hour increases on
both CVN 76 and CVN 77. When material did not arrive on time, the
shipbuilder tried to work around the missing item in order to remain on
schedule-which is less efficient than had the material been available when
planned. On CVN 77, for example, parts for a critical piping system were
delivered over a year late, necessitating work-arounds and resequencing of
work, driving labor costs up.
Other reasons for labor hour increases on CVN 76 and CVN 77:
CVN 76
o A 4-month strike in 1999 led to employee shortages in key trades,
contributing to a loss of learning with many employees not returning to
the shipyard. According to Navy officials, the shipbuilder was given $51
million to offset the strike's impact.
CVN 77
o Program schedule required concurrent design, planning, material
procurement, and production activities. Additional labor hours were spent
responding to design changes, which ultimately affected CVN 77 cost and
schedule.
o Due to unavailability of large-sized steel plates the shipbuilder had
to replan the ship's structure so it could be constructed with
smaller-sized plates. This required not only extensive redesign, but
resulted in additional production hours because laborers needed additional
time to fit and weld the smaller plates together.
Overhead and Labor Rates
While the total overhead and labor rate costs on both the CVN 76 and CVN
77 grew by $232 million over the life of the contract, labor hour
increases accounted for over half of that amount (See table 6.) According
to Navy officials, some of the overhead cost growth on CVN 76 can be
attributed to three major accounting changes since the contract was
awarded in late 1994. While these accounting changes increased overhead
costs, they resulted in a reduction of material costs. According to the
shipbuilder, overhead cost increases on CVN 77 can be attributed to
increases in pension and healthcare costs. Changes in the shipyard's
workload and employee benefit costs also led to overhead cost increases on
CVN 77. After delays in signing contracts for a carrier overhaul and the
next generation aircraft carrier, overhead costs had to be absorbed by the
CVN 77 program.
Table 24: Growth in Overhead Costs and Labor Rates
Dollars in millions
Analysis based on July 2003 (CVN 76) and March 2004 (CVN 77) data
Increase in Increase in Overhead cost as a
overhead overhead
Shipbuilder related to related to percent of total
reported growth in overhead and contract
Case study overhead growtha labor hours labor rates growth
ship
CVN 76 $263 144 $119
CVN 77 219 107 113
Total $482 251 $232
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures only costs associated with
overhead and labor rate changes. Increases in overhead related to growth
in labor hours are captured in the analysis of labor hour increases.
aContractor performance reports included $63 million in overhead costs for
CVN 76 and $40 million for CVN 77 that have been disallowed (not charged
to the government).
According to the shipbuilder, labor rate increases on CVN 76 resulted from
union negotiations following a strike at the shipyard, as well as
significant use of overtime labor, which is more expensive than normal
hourly wages. According to Navy officials, between 30 and 40 percent of
the work on CVN 76 was done on overtime in 2003.
Navy-Furnished Equipment
Navy-furnished equipment did not represent an area of cost growth on CVN
76 and CVN 77. On CVN 76, the costs for propulsion equipment decreased by
close to $145 million-driving down the overall cost of Navyfurnished
equipment. Since 2001, costs for Navy-furnished equipment on CVN 77,
however, have grown by $100 million. This growth on CVN 77 can be
attributed to increases in the cost associated with the Integrated Warfare
System-the carrier's combat system. The Integrated Warfare System included
new phased array radar that was being developed by the next generation
destroyer program. However, when the radar technology did not become
available as planned, the Navy decided to install a legacy system on the
ship. Because the shipbuilder was suppose to buy and install the
Integrated Warfare System as part of the original contract scope, the
costs for the Integrated Warfare System were removed from the contract and
used by the Navy to procure a legacy system as Navyfurnished equipment.
Appendix IV: San Antonio Class Amphibious Transport Dock Ship
Program Description The San Antonio class amphibious transport dock ship
is designed to transport Marines and their equipment and allow them to
land using helicopters, landing craft, and amphibious vehicles. The class
is expected to increase operational flexibility and survivability over
each ship's 40-year lifespan and to operate at lower cost than previous
amphibious transport ship classes. The new class is also designed to
reduce crew size and provide significant improvements in command, control,
communications, computer, intelligence, and quality of life.
In acquiring LPD 17, the lead ship in the class, a three-dimensional
computer-aided design tool and a shared data tracking system has been
used. The shared data tracking system was intended to provide significant
savings within the San Antonio class program through the reuse of critical
data in future design, construction, and operational activities. We
focused our review on the LPD 17 and 18.
Figure 8: San Antonio Class Amphibious Transport Dock Ship
Source: Navy.
Table 25: Major Events in the Acquisition of LPD 17 and LPD 18
February 1996 o Citing industrial base concerns and LPD 17's improved
survivability features, Congress authorized the LPD 17, accelerating the
Navy's schedule by 2 years.
March 1996 o Initial estimate developed for the class in support of a
milestone review. In developing the estimate, cost analysts used data from
the LHD and LSD amphibious ships, which were constructed in the 1980s. The
LHD is larger than the LPD, and the LSD is less technologically complex.
Cost analysts assumed that technologies would mature on schedule and that
acquisition reforms would produce savings.
December 1996 o LPD 17 cost-plus award fee contract awarded after a
competitive selection of the Avondale Alliance for detail design and
construction of LPD 17. The contract included options for construction of
LPD 18 and 19. Target costs were set for LPD 17 and LPD 18 at $644 million
and $391 million, respectively.
December 1998 o Contract modified to exercise option for construction of
LPD 18.
August 1999 o Litton Shipbuilding purchased LPD 17's prime contractor,
Avondale Industries.
December 1999 o Design schedule delays cause a 10-month slip in
anticipated delivery of LPD 17.
August 2000 o LPD 17 construction begins.
February 2001 o The Navy and Litton Alliance reassess the lead ship
construction schedule and delay LPD 17 delivery another 14 months to
November 2004.
April 2001 o Northrop Grumman Ship Systems assumed responsibility as
primary contractor for the LPD 17 program through an acquisition that
included Avondale.
September 2001 o Cost growth led to renegotiation of the contract.
September 2001 o The contract was converted to cost-plus incentive fee
contract. For LPD 17, the original award fee was based on the total cost
of the ship over its operational lifetime. The incentive fee contract tied
the fee to controlling construction costs. This shifted the focus of the
program from lowering future maintenance costs to delivering the ship.
November 2001 o Cost growth by more than 43 percent triggered a
Nunn-McCurdy unit cost breach, causing a new baseline to be established in
June 2002 and requiring $1.4 billion in additional funding.
February 2002 o LPD 18 construction begins.
June 2002 o With the Navy's approval, Northrop Grumman Ship Systems and
Bath Iron Works agreed to a swap that shifts all LPD construction,
including LPD 19, to Northrop Grumman and all future DDG 51 ships to Bath
Iron Works.
February 2005 o Navy requests $25 million in additional prior year
completion funding for LPD 18.
May 2005 o LPD 17 expected delivery date.
September 2005 o LPD 18 expected delivery.
Cost Experience on
LPD 17 and LPD 18
Sources: Navy, shipbuilder (data); GAO (presentation).
Budgets for the two LPD case study ships have grown by $1 billion, and
funds have been appropriated to cover these increases. However, the Navy
could need additional appropriations of $200 million to $300 million to
fund projected cost growth.
For detail design and construction of LPD 17, the Congress initially
appropriated $953.7 million to fund the construction contract (the basic
contract plus a budget for future changes) and acquisition of Navy-
furnished equipment. The Congress later appropriated $762 million to fund
LPD 18 construction. (See table 26.)
Table 26: Growth in Program Budgets for Case Study Ships
Dollars in millions
Initial and FY2005 President's budget Difference in budgets Case study ship
Initial FY2005 Total difference
Difference due to Navy-furnished equipment Difference due to construction
costsa
LPD 17 $954 $1,758 $804 $21 $784
LPD 18 762 1,011 249 3
Total $1,716 $2,769 $1,053 $24 $1,030
Sources: Navy (data); GAO (presentation).
aPart of the increased cost is due to changes in the scope of the
contract.
Since that time, the Congress has appropriated $1 billion to cover the
increases in the ships' costs. However, more funds will likely be needed
to cover additional cost growth for these two ships. We project that, if
the current schedule is maintained, total cost growth for the LPD 17 and
LPD 18 will exceed $1.2 billion and possibly reach $1.4 billion. (See
table 27.)
Table 27: GAO's Forecasts of Additional Cost Growth for Construction
Dollars in millions
Analysis based on data through May 2004
Percent of ship Amount already requested GAO's forecast for
GAO's forecast of total cost growth
Case study ship construction completed to cover increased cost additional cost
growth
LPD 17 95% $784 $112-$197 $896-$981
LPD 18 67 246 102-136 348-382
Total growth $1,030 $214-333 $1,244-1363
Sources: Navy, Shipbuilder (data); GAO (analysis).
Note: Cost growth is measured from original contract price, not from the
current contract target price. Forecast reflects expected price to the
Navy.
Our cost growth estimates-both low and high-are likely understated because
we assumed that the shipyards will maintain their current efficiency
through the end of their contracts and meet scheduled milestones. LPD 17
did not meet the planned December 2004 delivery date. Delivery is now
scheduled for May 2005, increasing the final cost of the ship.
Main Drivers of Cost Increases in labor hour and material costs account
for 76 percent of the
cost growth on LPD 17 and LPD 18 construction contracts. Navy-furnished
Growth for LPD 17 equipment-including radars, propulsion equipment, and
weapon and LPD 18 systems-represents just 2 percent of the cost growth.
The remaining
22 percent was due to increases in overhead and labor rates. (See fig. 9.)
Figure 9: Average Sources of Cost Growth on LPD 17 and LPD 18
Labor hours
Material
Overhead and labor rates
Navy-furnished equipment
Engineering costs (classified as material costs) associated with use of a
three-dimensional product model to design LPD 17 were a key contributor to
material cost growth. The design tool was not fully developed and
subsequent problems affected all aspects of the design. Subcontracts for
engineering design doubled, accounting for $215 million in cost growth on
LPD 17. Development of an integrated production data environment,
originally assumed to be funded by the state, has instead been shifted to
the contract, representing an additional $35 million in cost spread across
LPD. (See table 28.)
37
39
0 102030405060
Percent of cost growth
Percentage of overall cost growth due to shipbuilder construction costs
Percentage of overall cost growth due to cost of Navy-furnished equipment
Source: Navy (data); GAO (presentation).
The shipbuilder cited a number of direct causes for the labor hour,
material, and overhead cost growth in the two case study ships. The most
common causes were related to the concurrent development of a new and
unproven design tool and design of the lead ship, initial focus on
controlling total lifetime costs, and changes in employee pay and
benefits.
Materials
Table 28: Growth in Material Costs
Dollars in millions
Analysis based on data through May 2004
Material cost as a percent
of
Case study Increased material Percent increase total contract growth
ship costs
LPD 17 $400 103% 47%
LPD 18 93 39
Total $493
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events.
Labor Hours
Labor hours, the second largest component of cost growth, increased
significantly for the LPD 17 and LPD 18. For example, engineering labor
hours for the LPD 17 increased by over 100 percent from the original
proposal.
As the number of hours to construct the ship increased, total labor costs
grew, with the shipbuilder paying for additional overhead costs and
employee wages. We separated the overhead and labor rates associated only
with the additional hours and added this to the shipbuilder's reported
labor cost growth. (See table 29.) Our analysis captures all cost growth
associated with labor-including labor hours, overhead, and labor rates.
Table 29: Growth in Labor Hour Costs
Dollars in millions
Analysis based on data through May 2004
Overhead and labor rate Labor hour cost
as a
Shipbuilder reported costs Total cost due to percent of total
for increased labor contract
Case study labor cost growth hours increased labor growth
ship hours
LPD 17 a $182 $102 $284 33%
LPD 18 117 67 184 48
Total $299 $169 $468
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures all costs associated with
labor hour growth, including overhead and labor rates.
aLPD 17 relied heavily on subcontracted labor to design the ship. Since
these costs are captured as material, we did not include them in our
analysis of labor cost increases.
Factory inefficiencies and loss of skilled laborers, including significant
employee attrition (35 percent annually) contributed significantly to
labor hour increases. Difficulties with the design tool and turnover in
engineering staff led to increases in engineering labor hours and delayed
achieving a stable design. Without a stable design, work was often delayed
from early in the building cycle to later, during integration of the hull.
Shipbuilders stated that doing the work at this stage could cost up to
five times the original cost. On LPD 17, 1.3 million labor hours were
moved from the build phase to the integration phase. Consequently, LPD 17
took much longer to construct than originally estimated. Moreover, a
diminished workforce at Avondale required the busing of shipyard workers
from Ingalls Shipyard in Pascagoula, Mississippi to Avondale in New
Orleans, Louisiana and the subcontracting of skilled labor.
Program Overhead and While the total overhead costs on both the LPD 17 and
18 grew by
Labor Rates $0.5 billion over the life of the contract, labor hour
increases contributed to about half of that amount. (See table 30.)
Table 30: Growth in Overhead Costs and Labor Rates
Dollars in millions
Analysis based on data through May 2004
Portion Overhead cost rate
related to
Shipbuilder Portion overhead and
reported related to labor increases as a percent
Case study overhead growth growth in rates of total contract
ship labor hours growth
LPD 17 $277 $102 $175 20%
LPD 18 177 67 110
Total $454 $169 $285
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures only costs associated with
overhead and labor rate changes. Increases in overhead related to growth
in labor hours are captured in the analysis of labor hour increases.
According to Northrop Grumman, increases in overhead costs not related to
labor hour growth can be attributed largely to changes in the shipyard's
workload and employee benefit costs. Beginning in 2001, the shipyard
experienced a rise in overhead rates. For example, the overhead rates in
the 2004 latest estimate by Northrop Grumman are 39 percent higher than
what was originally proposed on the LPD 17 in 1996. Several factors helped
to increase overhead. For example; due to the loss of the bulk military
cargo T-AKE ship, the cancellation of the construction of a
commercial ship (American Classics Voyage), and the delay in signing the
contract for the next generation destroyer, overhead costs had to be
absorbed by the remaining contracts at the yard, including LPD. This led
to 36 percent of the increase in overhead rates-24 percent for the T-AKE
and cruise ship and 12 percent for DD(X).
According to the shipyard, changes in the financial market affected the
pension fund and the rise in medical care costs were responsible for 16
percent of the increase in the shipyards overhead rates.
Labor rates rose due to inflation impacts of an over 2-year delay in lead
ship delivery and subsequent changes in the procurement schedule and wage
rates negotiated with labor unions.
Navy-Furnished According to program officials, cost growth for
Navy-furnished equipment
Equipment on the LPD 17 was due to increased costs for a shock wave test
that was not anticipated in the original cost estimate. This cost was a
one-time increase, affecting only LPD 17 costs.
Appendix V: Virginia Class Submarine
Program Description The Virginia-class attack submarine, the newest class
of nuclear submarines, is designed to combat enemy submarine and surface
ships, fire cruise missiles at land targets, and provide improved
surveillance and special operations support to enhance littoral warfare.
Because the Virginia class is designed to be smaller than the Seawolf and
slightly larger than the Los Angeles class submarines-ships the new class
will eventually replace-the Virginia class is better suited for conducting
shallow-water operations. Major features of this new class of submarine
include new acoustic, visual, and electronic systems for enhanced stealth.
An objective of Virginia class is to reduce the life-cycle cost through
better design and engineering resulting in one third fewer man-hours than
were needed to construct Seawolf (SSN 21), the lead ship in the previous
class of attack submarines. The first ship, the SSN 774, was delivered in
October 2004. Our review focused on the SSN 774 and 775.
Figure 10: Virginia Class Submarine
Source: Navy.
Table 31: Major Events in the Acquisition of SSN 774 and SSN 775
October 1991 o Program initiated with focus on building more versatile
and less costly submarines.
July 1994 o Initial estimate developed for Virginia class based on
historical data from Seawolf (SSN 21) and Los Angeles (SSN 688) classes. A
sole source contract with Electric Boat was planned. Major challenges
involved estimating new technologies still under development and
estimating the cost impact of a 6-year gap in submarine production.
January 1996 o Contract for $1.4 billion awarded to Electric Boat for
detail design of Virginia class. Approximately 3.4 million man-hours for
one-time production start up activities were included.
February 1997 o Cost analysts updated estimates to reflect proposed
shipbuilder teaming agreement between Electric Boat and Newport News
Shipbuilding assuming teaming would be less expensive than dual sources
due to shipbuilder collaboration. Costs were increased to reflect
additional non-recurring effort for Newport News to reconstitute submarine
production.
February 1998 o Based on congressional direction that teaming agreement
would be the most efficient way to produce submarines in a low rate
production environment, the Navy authorized a teaming agreement between
Electric Boat and Newport News Shipbuilding. Deliveries of the ships would
alternate between shipyards with Electric Boat delivering the first ship.
According to the Navy, this change increased the estimated cost of
developing and building 30 submarines when compared to building them in a
single yard.
September 1998 o Contract is modified by $1.028 billion to fund
construction of Electric Boat's SSN 774 lead ship. Options for
construction of SSN 775 - 777 were included.
October 1998 o Construction of SSN 774 begins.
December 1998 o Contract is modified by $1.084 billion to initiate
construction of Newport News Shipbuilding's SSN 775 lead ship.
September 1999 o Construction of SSN 775 begins.
February 2001 o Navy requests $119 million for Virginia class in prior
year funding to cover cost growth.
December 2001 o Northrop Grumman Corporation acquires Newport News
Shipbuilding creating Northrop Grumman Newport News.
February 2002 o Prior year completion request of $227 million for
Virginia class.
April 2003 o Prior year completion request of $327 million for Virginia class
triggers a Nunn-McCurdy unit cost breach. June 2004 o Planned delivery date
for SSN 774.
October 2004 o SSN 774 delivered 4 months late, an improvement over the
Seawolf and Los Angeles lead ships which were delivered 25 and 26 months
late, respectively.
February 2005 o Fiscal Year 2006 President's Budget recognizes an
increase in the budget of $82 million for SSN 774 and $223 million for SSN
775. These increases are funded through transfer of funds and prior year
funding of $28 million for SSN 774 and $97 million for SSN 775.
June 2005 o Initial planned delivery date for SSN 775.
March 2006 o Current planned delivery date for SSN 775.
Sources: Navy, shipbuilder (data); GAO (presentation).
Cost Experience on The Fiscal Year 2005 President's Budget showed that
budgets for the two Virginia class case study ships have increased by $734
million. However,SSN 774 and SSN 775 based on data of July 2004, we
projected additional cost growth on contracts for the two ships is likely
to reach $840 million and could be
higher. In fiscal year 2006 budget, the Navy has requested funds to cover
cost increases that are now expected to reach approximately $1 billion.
The fiscal year 2005 budget for the SSN 774 and SSN 775 is about $6.2
billion, compared with the initial fiscal year 1998 budget request of $5.5
billion. (See table 32.) Ship construction costs comprise the majority of
this increase.
Table 32: Growth in Program Budgets for Case Study Ships
Dollars in millions
Initial and FY2005 President's budget Difference in budgets Case study ship
Initial FY2005 Total difference
Difference due to Navy-furnished equipment Difference due to construction
costsa
SSN 774 $3,260 $3,682 $422 $95 $327
SSN 775 2,192 2,504 312 18
Total $5,452 $6,186 $734 $113 $621
Sources: Navy (data); GAO (presentation).
aPart of the increased cost is due to changes in the scope of the
contract.
While the Congress has appropriated funds to cover the increases in the
ships' costs, more funds will be needed to cover additional cost growth
for these two ships. In its fiscal year 2006 budget submission the Navy is
requesting an additional $125 million in prior year completion funding
between fiscal years 2006 to 2007 for the case study ships. We calculated
a range of the potential growth for the two case study ships and found
that the total projected cost growth is likely to exceed $724 million and
could reach $840 million or higher. (See table 33.)
Table 33: GAO's Forecasts of Additional Cost Growth for Construction
Dollars in millions Analysis based on July 2004 data
Percent of ship Amount already
requested
to cover GAO's forecast GAO's
construction contractor's for forecast of
total
Case study completed increased cost additional cost cost growth
ship growth
SSN 774 Delivered $327a $0-0 $327-327
SSN 775 88% 294 103-219 397-513
Total $621 $103-219 $724-840
growth
Main Drivers of Cost Growth for SSN 774 and SSN 775
Sources: Navy (data); GAO (analysis).
aProgram officials indicated that $70 million in additional funding has
been requested for SSN 774 completion.
Our cost growth estimates-both low and high-may be understated because we
assumed that the shipbuilders will maintain their current efficiency
through the end of their contracts and meet scheduled milestones. Any
slips in efficiency and schedules would likely result in added costs. For
example, the delivery date for SSN 775 is expected to slip by as many as
8.5 months, which could increase the final cost of the ship.
Our analysis shows that the submarine contract costs have grown because
initial construction costs were underestimated-especially the costs
associated with the cost of material and number of labor hours needed to
build the ships. For the two case study ships we examined, we found that
increases in the number of labor hours and material costs to build the
submarines accounted for 83 percent of the cost growth on shipbuilding
construction contracts. Navy-furnished equipment, including radars,
propulsion equipment, and weapon systems, caused 14 percent of the cost
growth. We found that ship overhead-such as employee benefits and shipyard
support costs-and labor rate increases accounted for 3 percent of cost
growth.
Figure 11: Average Sources of Cost Growth on SSN 774 and SSN 775
Labor hours
Material
Overhead and labor rates
Navy-furnished equipment
43
0 102030405060
Percent of cost growth
Percentage of overall cost growth due to shipbuilder construction costs
Percentage of overall cost growth due to cost of Navy-furnished equipment
Source: Navy (data); GAO (presentation).
In negotiating the contract for the first four Virginia class ships,
program officials stated they were constrained in negotiating the target
price to the amount funded for the program, thereby risking cost growth at
the outset. The shipbuilders said that they accepted a challenge to design
and construct these ships for $748 million less than their estimated costs
because the contract protected their financial risk. Despite the fact that
there was significant risk of cost growth, the Navy, based on guidance at
the time, did not identify any funding for probable cost growth.
We analyzed shipbuilder contract costs to identify the sources of cost
growth. Using shipbuilder cost data, we allocated the sources of
shipbuilder cost growth on the contract into three categories-labor hours;
material costs; and labor and overhead rates. Since labor costs and
overhead costs can change due to labor hours, labor rates, and rates
associated with individual elements of overhead-or a combination of
these-we examined each in isolation by separating the program overhead
cost associated with an increase in labor hours from costs that resulted
from an increase in overhead rates, such as an increase in health care
costs.
Materials Due to high risk that specialized material could not be
procured for the amount budgeted, the Navy agreed to purchase this
material as a cost plus fixed fee item. This agreement protected the
shipbuilder from having to
fund any resulting cost increases for highly specialized material. Indeed,
cost growth for material increased by $350 million for the two Virginia
class submarines we examined.
Table 34: Growth in Material Costs
Dollars in millions Analysis based on July 2004 data
Total dollars Material cost
due to increased as a percent of
Case study ship material costs Percent increase total contract growth
SSN 774 $141 43
SSN 775 209 56
Total $350 99
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events.
The Navy and shipbuilders attribute material cost growth to several
factors including
o unrealistic budgets not supported by current vendor costs,
o diminished supplier base for highly specialized materials,
o nonrecurring costs for Computer Data Integration between shipbuilder
teams,
o lack of design maturity for certain electronic components, and
o full funding of ships in the year of authorization.
Shipbuilders stated they based more than 70 percent of their estimate for
major material costs on updated vendor quotes while the Navy relied on
historical costs that were not analogous to the low number of submarines
being planned for construction. While the Navy knew there would be a price
penalty for a 6-year gap in submarine production, there were no studies or
actual data to support what the overall effect would be. Thus, Navy cost
estimators assumed that costs for major material items would increase by
20 percent. When the Navy negotiated the costs for Virginiaclass high
value, specialized material, the shipbuilder agreed to take on the
challenge of achieving lower costs in exchange for funding these materials
on a cost-plus-fixed-fee basis.1 By the time the lead ship was delivered
1 Under this arrangement, the Navy is responsible for any cost growth.
8 years later, the true cost increase for highly specialized material was
closer to 60 percent more than historical costs.
Following the cancellation of the prior submarine program-Seawolf- and a
decrease in submarine production of three to four submarines per year to
one over a period of 6 years, many vendors left the nuclear submarine
business and focused instead on more lucrative commercial product
development. As a result, prices for highly specialized material increased
due to less competition and a lack of business. For example, many vendors
were reluctant to support the Virginia class submarine contract because
costs associated with producing small quantities of highly specialized
materials were not considered worth the investment- especially for
equipment with no other military or commercial applications.
Material costs also increased due to nonrecurring costs for integrating
computer data so that the shipbuilders could work from a common design. In
addition, costs to develop high-risk systems like the array and exterior
communication system were underestimated. Recognizing the significant cost
risk involved, the Navy procured these systems under a separate contract
line item that guaranteed the shipbuilders a fixed fee and made the Navy
responsible for funding all cost growth.
Finally, the Navy believes that the block-buy contract has contributed to
increased material costs. Under a block-buy contract, subcontracts for
submarine materials are for single ships spread over several years.
According to the Navy, this type of acquisition approach does not take
advantage of bulk-buy savings and incurs the risk that funding will not be
available in time to order the material when needed. In addition, since
ships are funded individually, the Navy believes suppliers are unwilling
to risk investing in technology improvements due to the uncertainty that
future ships will not be purchased. To stabilize the vendor base, the Navy
awarded a multiyear contract that commits the Navy to purchasing
additional submarines. While a multiyear contract can provide such
savings, a program must meet criteria to demonstrate a sufficient level of
stability for such a contract. In June 2003, we noted several aspects of
the Virginia class program that indicated instability.2 Another factor to
be considered in using multiyear contracts is the budget flexibility the
2 GAO-03-895R, Multiyear Procurement Authority for the Virginia Class
Submarine Program, (Washington, D.C.: June 23, 2003).
government gives up in exchange for the commitment of funds for the future
years of the contract.
Labor Hours Labor cost increases have led to $339 million in cost growth
for the SSN 774 and SSN 775 combined. Problems with mastering state-of-the
art design tools, first in class technical and teaming issues, and
material availability all contributed to the labor cost growth.
Table 35: Growth in Labor Hour Costs
Dollars in millions Analysis based on July 2004 data
Overhead and labor Labor hour cost
as a
Shipbuilder reported rate Total cost due to
costs for increased percent of total
Case study labor cost growth labor increased labor contract growth
ship hours hours
SSN 774 $149 $10 $159 55%
SSN 775 218 (38) 180
Total $367 $(28) $339
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures all costs associated with
labor hour growth, including overhead and labor rates.
We found that SSN 774 required almost 3 million additional labor hours
than planned, reflecting a growth of 25 percent. (See fig. 12.) In
addition, we found that SSN 775 required almost 4 million more labor hours
than planned. Approximately 3.4 million nonrecurring labor hours for SSN
774 were procured on a separate contract line item and therefore not
included in our analysis while some SSN 775 nonrecurring labor hours are
embedded in the labor hours for that ship.
Figure 12: SSN 774 Lead Ship Labor Hour Growth
Thousands of hours 3,000
2,500
2,000
1,500
1,000 500 0
.Apr
lyJuOct.
Dec.
.Apr
lyJu
Sept.Dec.
.Mar
Sept.Dec. . Sept.Dec.
.Mar
lyJu
Sept.Dec.
.Apr
lyJu
Sept.
enJu
enMarJu
1999 2000 2001 2002 2003 2004
Source: Shipbuilder (data), GAO (presentation).
Technical issues commonly associated with first-in-class ships also
contributed to the overall labor cost growth. For example, shipbuilders
experienced problems with crossed hydraulic lines on the lead ship. In
addition, problems with the torpedo tube and weapons handling design
issues also contributed to labor hour growth in both ships. Labor hours
also increased as quality problems discovered for a component made by one
shipyard were reworked by the shipyard integrating the components. Because
the shipyard doing the integration was not as familiar with the effort,
the work was not completed as efficiently.
Late material deliveries also disrupted the work-flow sequence. Because
many vendors either went out of business or focused on developing new
commercial products in response to low demand, the Navy was no longer
considered a preferred customer. In cases where there was no ready
supplier, the shipbuilder had to request old subcontractors to supply the
highly specialized material. This caused delays in material deliveries as
well as quality problems arising from strict inspection processes that
subcontractors were no longer familiar with. Although the shipbuilders
tried to work around late material deliveries when they could, this caused
workers to perform less efficiently than had the material been available
when scheduled. Moreover, when the material did arrive, the shipbuilders
had to work overtime to make up the schedule causing additional growth in
labor costs.
Navy-Furnished Equipment
According to Navy program officials, radar costs increased due to more
design effort needed to fix problems associated with the Seawolf program.
Other costs increases were driven by changes in how certain items were
purchased. For example, the advanced display system was recently
established as a line item in the budget when in the past it was paid for
as part of the shipbuilder's construction contract. Moreover, the Navy
initially planned to use research and development funds to cover costs for
the propulsor but switched to ship construction funds instead, leading to
an increase in the program's budget for Navy-furnished equipment.
Ship Overhead and Labor Our analysis shows that program overhead costs and
increases in labor
Rates rates were not significant sources of cost growth-causing
approximately 3 percent of the cost growth. To isolate true increases in
overhead rates from increases that were a consequence of labor hour
increases, we separate the two in table 36.
Table 36: Growth in Overhead Costs and Labor Rates
Dollars in millions Analysis based on July 2004 data
Increase in Increase in Overhead cost as a
overhead overhead
Shipbuilder related to related to percent of total
reported growth in overhead and contract
Case study overhead growth labor hours labor rates growth
ship
SSN 774 $0 $10 ($10) (3%)
SSN 775 0 (38) 38
Total $0 $(28) $28
Sources: Shipbuilder (data); GAO (analysis).
Note: We compared initial target cost to the current estimate at
completion to determine total contract cost growth. Cost growth may be due
to Navy changes in contract scope, shipbuilder performance, or
unanticipated events. Our analysis captures only costs associated with
overhead and labor rate changes. Increases in overhead related to growth
in labor hours are captured in the analysis of labor hour increases.
Costs associated with growth in labor hours are shown in table 35
calculations.
According to the shipbuilder, overhead and labor rate increases were
related to pension, workers compensation, and health care costs rising
beyond what was expected. Furthermore, when other ship acquisitions did
not materialize, shipyard overhead costs were spread over a fewer number
of contracts causing an increase in the Virginia class overhead costs.
Similarly, the loss of business caused the shipbuilders to lay off skilled
workers. According to the shipbuilders, many of the experienced workers
did not return to the shipyard. Hiring and training new workers increased
costs.
We found that one shipbuilder was affected by labor rate increases.
Following a strike at the shipyard, union negotiations resulted in four
pay increases totaling an average of $3.10 per hour.
Appendix VI: GAO's Forecast of Additional
Costs to Complete Construction Contracts
This appendix discusses GAO's forecast of future cost growth for all ships
in construction that are more than 30 percent complete. The forecast is
also compared with the shipbuilders' forecasts of estimated costs at
completion.
Figure 13: Comparison of Shipbuilders' and GAO's Forecasts of Additional
Construction Costs for Six Classes of Ships Actively under Construction
Based on data available to GAO through July 2004
Ships under construction
Percent of
ship constuction completed
Shipbuilder's estimate of costs to complete ship
Low end of GAO's forecast of costs to completea High end of GAO's forecast
of costs to completeb
DDG CVN LPD SSN T-AKE LHD
Sources: Navy, Shipbuilder (data); GAO (analysis).
Notes: Active construction in this table means ships are at least 30
percent complete. Cost growth figures are in millions of dollars. Ships'
names that are bolded are case study ships.
aWe based the lower end of our cost forecast range on the costs spent to
date added to the forecast cost of work remaining. The remaining work was
forecasted using the cumulative cost performance index efficiency factor.
Studies have shown that using this method is a reasonable estimate of the
lower bound of the final cost.
bFor the upper end of our cost range, we relied on either the actual costs
spent to date added to the forecast of remaining work with an average
monthly cost and schedule performance index or a cost/percent complete
trend analysis, whichever was higher.
o CVN 76 and CVN 77: CVN 76 was delivered to the Navy in 2003. While we
forecasted an overrun of up to $586 million over the initial target price
for CVN 77, the fiscal year 2006 budget request indicates a need of $870
million in prior year funding.
o SSN 774-SSN 777: SSN 774 was delivered to the Navy in October 2004. We
found that the contractors' forecasts are unlikely to be achieved based on
continuing cost growth on the remaining 3 ships. In addition, the SSN 776
and SSN 777 are the follow-on ships to a new class and still may
experience production problems that could lead to future cost growth.
o DDG 91-DDG 101: The DDGs have experienced cost growth at both
shipyards. All the DDGs under construction at Bath Iron Works and more
than 30% complete have experienced cost growth. Similarly, cost growth is
also expected on the DDGs built by Northrop Grumman.
o LPD 17-LPD 20: LPDs currently under construction are likely to
experience significant cost overruns. On all of the LPDs, with the
exception of LPD 18, the shipbuilder is estimating overall cost growth to
be at the lower end of our predicted range. Hence we believe, the
shipbuilder's forecast of cost growth is optimistic.
o T-AKE: Major cost growth is being predicted for T-AKE 1. We estimate
that costs could grow more than $70 million beyond the initial contract
price. The shipbuilder believes that escalating material costs resulting
from rising commodity prices and unfinalized vendor subcontracts are
driving contract cost growth.
o LHD 8: It also has the potential for significant cost growth-as much
as $177 million more than what was anticipated. Cost growth thus far is
attributed to increases in overhead and general and administrative costs.
Appendix VII: Comments from the Department of Defense
Appendix VIII: GAO Contacts and Staff Acknowledgments
GAO Contacts Acknowledgments
(120237)
Paul L. Francis (202) 512-2811 Karen Zuckerstein (202) 512-6785
In addition to the contacts named above, Margaret B. McDavid, Christina
Connelly, Diana Dinkelacker, Christopher R. Durbin, Jennifer Echard, R.
Gaines Hensley, Ricardo Marquez, Christopher R. Miller, Madhav Panwar,
Karen Richey, Karen Sloan, Lily Chin, and Marie Ahearn made key
contributions to this report.
GAO's Mission The Government Accountability Office, the audit, evaluation
and investigative arm of Congress, exists to support Congress in meeting
its constitutional responsibilities and to help improve the performance
and accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.
Obtaining Copies of The fastest and easiest way to obtain copies of GAO
documents at no cost
is through GAO's Web site (www.gao.gov). Each weekday, GAO postsGAO
Reports and newly released reports, testimony, and correspondence on its
Web site. To Testimony have GAO e-mail you a list of newly posted products
every afternoon, go to
www.gao.gov and select "Subscribe to Updates."
Order by Mail or Phone The first copy of each printed report is free.
Additional copies are $2 each. A check or money order should be made out
to the Superintendent of Documents. GAO also accepts VISA and Mastercard.
Orders for 100 or more copies mailed to a single address are discounted 25
percent. Orders should be sent to:
U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548
To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061
To Report Fraud, Contact:
Waste, and Abuse in Web site: www.gao.gov/fraudnet/fraudnet.htm
E-mail: [email protected] Programs Automated answering system: (800)
424-5454 or (202) 512-7470
Congressional Gloria Jarmon, Managing Director, [email protected] (202)
512-4400 U.S. Government Accountability Office, 441 G Street NW, Room 7125
Relations Washington, D.C. 20548
Public Affairs Paul Anderson, Managing Director, [email protected] (202)
512-4800 U.S. Government Accountability Office, 441 G Street NW, Room 7149
Washington, D.C. 20548
*** End of document. ***