93doc04 0

CBO
PAPERS
ALTERNATIVES FOR
THE U.S. TANK
INDUSTRIAL BASE
February 1993
CONGRESSIONAL BUDGET OFFICE

CBO
PAPERS
ALTERNATIVES FOR
THE U.S. TANK
INDUSTRIAL BASE
February 1993
CONGRESSIONAL BUDGET OFFICE

SECOND AND D STREETS, S.W.
WASHINGTON, D.C. 20515
NOTES
All costs are expressed in 1993 dollars of budget authority.
All years, unless otherwise noted, are fiscal years.
Numbers in tables may not add to totals because of rounding.

PREFACE
The U.S. military has no plans to purchase new tanks after 1992. Without
U.S. sales, tank production in this country might cease, leaving in question the
ability of U.S. tank manufacturers to produce tanks should they be needed
during a crisis. This paper, prepared at the request of the House Committee
on Armed Services, explores various options for maintaining the U.S. tank
industrial base. The options differ widely in terms of their cost and the
insurance they provide against an unforeseen need for new U.S. tanks. This
information may be useful to the Congress as it debates the fate of the
defense industrial base in this time of fiscal constraint. In keeping with the
Congressional Budget Office's (CBO's) mandate to provide objective analysis,
the paper makes no recommendations.
Frances M. Lussier prepared this paper under the general supervision

of Robert F. Hale and R. William Thomas. William P. Meyers of CBO's
Budget Analysis Division provided cost analyses. The author wishes to thank
Wayne Glass for his assistance, as well as Sherry Snyder, who edited the
report, and Cynthia Cleveland, who prepared it for publication.
Robert D. Reischauer
Director
February 1993
CONTENTS
SUMMARY 1
INTRODUCTION AND BACKGROUND 8
The Tank Industrial Base 8

The Army's Tank Requirements 10
Implications for the Tank Industrial Base 19
OPTIONS FOR MAINTAINING

THE TANK INDUSTRIAL BASE 20
Alternative I: Lay Away the Tank Industrial Base 20

Alternative II: Maintain Production, but at a
Low Rate 25
Alternative III: Convert Older Abrams Tanks
to Newer Models 26
COMPARING THE ALTERNATIVES 31
Arguments in Favor of Layaway 31

Arguments in Favor of Low-Rate Production
or Conversion 34
CONCLUSION 36
ALTERNATIVES FOR THE U.S. TANK INDUSTRIAL BASE
TABLES
S-l. Costs of Alternatives for Preserving the

Tank Industrial Base 3
S-2. Annual Costs Through 1998 for

Various Alternatives 5
1. Overview of the U.S. Tank Industrial Base 9
2. Funds Spent per M1A1 Tank in Various States 12
3. The U.S. Army Tank Fleet 15
4. Annual Costs Through 1998 for Various

Alternatives 23
5. Costs from 1994 Through 1998 for Alternative

I (Lay Away the Tank Line), by Function
and Facility 24
6. Characteristics of the Abrams Tank Inventory 28
7. Costs Through 1998 for Alternative III

(Convert Mis to MlA2s) 30
8. Costs and Restart Times of Various Alternatives 32
9. Effects of Various Alternatives on the Active

Inventory and Capability of Tanks in 2000 33
FIGURES
1. Distribution of M1A1 Production Funds 11
2. U.S. Army Total Tank Inventory 14
3. Comparison of Various Tank Fleets 17

SUMMARY
The shrinking of the U.S. military, coupled with the disappearance of a long-
time foe and the unprecedented peacetime investment in modern weapons
that occurred in the 1980s, has led to sharp reductions in the planned
purchases of military weapons. In particular, the Bush Administration did not
plan to buy any new tanks for U.S. forces for the next 15 or more years.
Without U.S. purchases, the U.S. industrial base dedicated to producing tanks
could close and would eventually atrophy.
The tank industrial base in this country involves facilities operated by

several major contractors that are supported by some 18,000 subcontractors
and suppliers. Some skills and machines used in producing tanks would not
be maintained in the absence of ongoing tank production and could not be
regained quickly if lost as a result of disuse. Uncertainty about future
requirements, together with the delays associated with reassembling the skilled
workers and lines needed to produce tanks, may argue for maintaining some
U.S. capacity to produce tanks even if no new tanks are needed now.
Maintaining such capacity by keeping the tank line open, however, would cost
a minimum of $250 million annually after 1995 and could cost up to three
times more.
The Congressional Budget Office (CBO) examined three ways of

preserving some or all of the components of the tank industrial base. The
first would retain no active components but would preserve the plants and
lines that are used exclusively to produce tanks or tank parts. The second
alternative represents the other extreme and would keep the entire industrial
base active by producing a small number of new tanks each year. The last
alternative would keep most of the base active by upgrading older models of
the Abrams Ml tank to the latest and most capable model, the M1A2. These
options would all add roughly the same amount of capability to the U.S. tank
fleet, but they vary greatly in cost and in the amount of insurance they would
provide against an unforeseen need for new tanks sometime in the future.
They would also differ greatly in the amount of employment they would
provide in the defense sector, but CBO did not analyze this aspect of the
issue.
Alternative I: Lay Away the Tank Industrial Base
With no further purchases of new tanks, one alternative would be to lay away
and preserve the physical portions of the tank industrial base after all orders
for tanks have been filled. Taking into account the Bush Administration's
plan submitted in February 1992, Congressional action on the fiscal year 1993
budget, and current orders, this layaway process could begin in 1994 and could
be completed in 1997 with the shutdown of the assembly line. Additional

foreign military sales could, however, delay the final layaway.
This plan, espoused by the Bush Administration as recently as April

1992, is consistent with the Army's lack of a requirement for new tanks. In
1995, the Army's inventory of 7,780 Abrams tanks will be sufficient to equip
all Army units that the Bush Administration planned to field, to provide war
reserves and prepositioned stocks, and to meet other support needs.
This required inventory of roughly 8,000 tanks would be more capable

than almost any other tank fleet in the world in 1995, except that which
Russia could field throughout its vast reaches. Based on a scoring system that
reflects both the quantity and quality of a country's weapons, the Abrams tank
fleet would be 12 times more capable than the total tank forces fielded by
postwar Iraq, North Korea, or Libya. The Abrams fleet would also be five
times more capable than the very large tank forces of China. Taken as a
whole, the Russian fleet would be twice as capable as the U.S. Abrams fleet.
When considered separately, however, the widely separated European and
Asian Russian fleets would be much less formidable. The Abrams fleet would
be almost 60 percent more capable than the relatively modern Russian tank
fleet stationed in Europe and about two-thirds as capable as Russia's less
sophisticated but larger tank fleet dispersed throughout its Asian region.
The Abrams tank fleet envisioned for 1995 will not only be very capable
but also relatively young; the oldest Abrams tank will be only 15 years old.
Given that the useful service life of a tank is typically assumed to be 30 years,
the United States should not need to produce new tanks to replace aging
vehicles until 2010 or later, again confirming the Bush Administration's plan
to buy no new tanks for at least 15 years.
Laying away the tank line would be the least costly alternative examined
in this paper, requiring a total of $340 million for the five years from 1994
through 1998 (see Summary Table 1). (All costs in this paper are expressed
in 1993 dollars of budget authority.) These funds would be used to preserve
the physical portions of the tank industrial base~the plants, machines, and
assembly lines~and for personnel costs associated with ceasing production of
tanks. Thereafter, maintaining facilities in mothballed status and providing
security would cost about $40 million a year. The Army estimates that a line
that has been laid away for more than two years would take about 72 months
and $1.1 billion to start producing tanks again at surge rates (defined as 120
tanks a month). Moreover, based on historical evidence, the first vehicles off
a restarted tank line might not be of high quality.
Although the time required to produce significant numbers of tanks

from a line that has been laid away far exceeds the length of any likely
contingency operation, that would also be true for the other two alternatives.
Unlike the other alternatives, however, the very long period of six years
required to reach surge rates from a laid-away base could be of concern if a
major new threat to U.S. security emerges, particularly if there is a significant
delay before this country recognizes the threat's existence and responds to it.
Of more concern, perhaps, is the immediate loss of the ability to build

or modernize tanks in this time of global uncertainty. In the aftermath of a
highly armored confrontation in the Persian Gulf War and the breakup of a
SUMMARY TABLE 1. COSTS OF ALTERNATIVES FOR

PRESERVING THE TANK INDUSTRIAL BASE
(In millions of 1993 dollars)
Cost to Reach
Annual Surge Caoacitvc
Cost, Cost After Time
1994-1998a 1998b (Months) Funds
Alternative I:
Lay Away Tank Line 340 40 72 1,110
Alternative II:
Produce at Low Rate 3,830d 740d 47 600
Alternative III:
Convert Older Models 2,210 d - e 650d-f 56 530
SOURCE: Congressional Budget Office based on Army data.
a. Does not include $15 million to lay away the assembly line at the Detroit Arsenal Tank Plant.
b. Assuming the program continues indefinitely.
c. Surge capacity is 120 tanks per month and would require two assembly lines operating on multiple shifts.
d. Assumes additional production of 120 tanks per year to fill orders from foreign governments.
e. The proposed budget revisions submitted by then Secretary of Defense Dick Cheney in January 1993 include
an estimate of $720 million as a minimum sustaining level.
f. According to the proposed budget revisions submitted in January 1993, a minimum annual investment of $250
million would sustain the program, but would convert only 60 tanks per year rather than the 198 funded in
the alternative.
45-year empire in Eastern Europe, this may not be an auspicious time for the
United States to be without the capability to produce tanks. Continuing to
maintain the tank industrial base, at least until the new world order becomes
more certain, might be a sound insurance policy.
Alternative II: Maintain Production, but at a Low Rate
To reduce both the risk of losing the ability to make tanks in the near term
and the time needed to start producing tanks in large numbers at some time
in the future, should they be needed, the United States could keep the entire
industrial base active by producing new tanks at a low rate of 10 per month.
Maintaining production at a low rate would ensure that workers trained in
each of the skills critical to building tanks would be available at all times,
though not necessarily in large numbers. Keeping the base warm would
appreciably shorten the time needed to start producing new tanks at surge
rates, from as much as 72 months from layaway to 47 months from low-rate
production. Contractors would need this time to hire and train the additional
personnel required to operate two assembly lines on multiple shifts. Although
47 months would still be longer than the duration of the most probable future
contingencies, the shorter response time could be significant if a major new
threat emerges that would require an increase in the size of the Abrams tank
fleet.
This alternative would also provide the Army with 120 additional new
tanks every year-tanks that the Army does not need to support the force
planned for 1995. Because they would be a more recent and capable version,
newly produced tanks could replace some of the older Abrams tanks in the
inventory. The overall improvement in capability resulting from new tanks
delivered to the Army by the year 2000 would be small, however~on the order
of 2 percent.
The major disadvantage of this alternative is its high cost, which could
total $3.8 billion from 1994 through 1998~more than 10 times the cost of
laying away the line (see Summary Table 2). In 1998 and thereafter, annual
costs to produce 120 tanks each year would be more than $700 million,
compared with $40 million for the previous alternative.
Alternative III: Convert Older Abrams Tanks to Newer Models
As a compromise between producing tanks at a low rate and laying away the
tank industrial base, the United States could convert older and less capable
models of the Abrams tank to the newest, most sophisticated configuration.
This latest model, designated the M1A2, has a bigger, more powerful cannon
than the Abrams tanks produced in the early 1980s, as well as better armor
protection, digital electronics, better electro-optics, and a system for
preventing harmful nuclear, biological, or chemical contaminants from
reaching the crew. Converting an original Ml to an M1A2 involves taking the
tank almost totally apart and replacing all of the major components except for
the engine and transmission. Thus, conversion would involve all parts of the
industrial base except those that produce engines and transmissions. The
Army currently plans to fund a program to upgrade 210 Ml tanks from 1993
through 1995. The converted M1A2 tanks, in addition to new M1A2 tanks
already purchased, would equip one division in the Army's contingency force,
which would fight in regional wars such as Operation Desert Storm. To equip
the entire contingency force, the Army would need to upgrade a total of 1,002
SUMMARY TABLE 2. ANNUAL COSTS THROUGH 1998

FOR VARIOUS ALTERNATIVES
Five-
Year
1994 1995 1996 1997 1998 Total
Alternative I:
Lay Away Tank Line 50 30 140 80 40 340
Alternative II:
Produce at Low Rate* 780 800 760 740 750 3,830
Alternative III:
Convert Older Models
Phase I 80 130 0 0 0 210
Phase H a - b 0 150 550 650 650 2,000
SOURCE: Congressional Budget Office based on Army data and Congressional action on the 1993 budget
request.
NOTE: Costs do not include about S15 million to lay away the assembly line at the Detroit Arsenal Tank Plant.
a. Assumes annual sales to foreign militaries of 120 tanks.
b. According to then Secretary of Defense Dick Cheney's proposed budget revisions submitted in January 1993,
a minimum annual investment of $250 million would sustain the program after 1995, but at 60 conversions per
year rather than the 198 conversions funded in the alternative.
Ml tanks. Although the Army would like to convert an additional 792 tanks
to the M1A2 configuration, it is not yet sure how it will structure or fund such
a program.
However it is structured, a conversion program would avoid producing

new tanks that would only be surplus to those the Army needs to support the
forces proposed for 1995. By improving older models, a program that
upgraded a total of 1,002 tanks in the next seven years would increase the
capability of the entire required tank fleet by 3 percent by the year 2000, and
that of the contingency force by 10 percent. The conversion process could
also eventually lead to an Army inventory with fewer models of the Abrams
tank, which should reduce operating costs. Perhaps most important, however,
maintaining the unique components of the industrial base would provide some
insurance in this time of uncertainty.
By other measures, this approach lies between the previous two. About
56 months would be needed to start producing new tanks at surge rates from
a line dedicated to conversion. This period falls between the delays
associated with the other two options and is sufficiently long so that large
numbers of additional new tanks would not be available for short-notice
contingencies. The costs would also fall between those of the previous two
options. In 1994 and 1995, the Army plans to spend at least $220 million on
conversion, three times the cost of layaway but about one-seventh the cost of
low-rate production. Beyond 1995, costs would still fall between those of the
other two options, but the level of funding would depend on how the
conversion program is structured. Under the Army's plan as of fall 1992,
annual costs might total $650 million, which would sustain the tank line and
convert 198 tanks a year to the M1A2 configuration. Under the budget
revisions submitted by then Secretary of Defense Dick Cheney in January
1993, conversion costs might amount to as little as $250 million a year beyond
1995. This lower funding would sustain the tank line (assuming foreign sales
of 120 tanks a year), but the Army has indicated informally that it is not clear
how many tanks could be converted.
Should the United States Maintain Its Tank Industrial Base?
The Congress has directed that funds appropriated for fiscal year 1992 be
used in 1993 to convert older tanks to newer models. The question it must
now address is whether the conversion program should continue and, if so, at
what pace and for how long.
If the overriding criterion is to limit costs, then the option to lay away
the tank base should be pursued instead. In the next two years, 1994 and
1995, conversion would cost more than twice as much as layaway.

Subsequently, depending on the pace of conversion, upgrading tanks could
cost from as little as two to more than 10 times as much as layaway (see
Summary Table 2). If the conversion program continues at a high pace after
1995, the United States could spend more in two years to convert tanks than
it would to reactivate the base and attain a surge rate of production.
Closing down and laying away the tank line, especially during the next
two years, involves risks that a program that sustains the tank base would
avoid. Inability to respond to an unforeseen threat to U.S. security is one
such risk. Russia could again become a highly militarized society, but that
seems unlikely. The emergence of a developing nation equipped with
sophisticated tanks obtained on the global arms market is perhaps a more
likely scenario. U.S. responses to such threats could require additional tanks
or upgraded versions of the Abrams tank in order to ensure that U.S. forces
retain their superiority.
The time and effort needed to resurrect a tank line that has been laid
away are also uncertain. If restarting production takes longer than is currently
estimated, the risk of failure to meet an unanticipated threat is greater. Risks
also increase if the ability of a reopened line to produce quality tanks has
been overestimated. The risks associated with restarting the line may be
greater than currently estimated because the planned shutdown is likely to be
at least 15 years, a rather long time. A conversion program, or one that
maintains low-rate production, would ensure the availability of quality
components as long as it continued. Programs that maintained most of the
base in active status might avoid the problems of quality associated with
producing tanks from a restarted production line that had been long idle.
Even if these programs were continued for only a few years, they would also
provide additional time to study the issues related to closing down the tank
line completely.
In conclusion, reducing risks by maintaining most of the tank base after

1995 through a conversion program could cost between $250 million and $650
million a year. The costs would represent a small fraction of the Army's total
budget and would provide some insurance against unforeseen threats and
risks. Such insurance might be worth buying for a few years until the trends
in the former Soviet Union and other parts of the world become clearer. In
the years to come, however, the risks incurred by ending support for the tank
industrial base should continue to be weighed against the substantial cost of
maintaining it.
INTRODUCTION AND BACKGROUND
During the 1980s, the Army embarked on an ambitious modernization

program, purchasing-among many other things-more than 7,000 new tanks,
more than 5,500 new fighting vehicles, 728 new attack helicopters, and 478
new multiple rocket launchers for a total cost of almost $45 billion. These
new lethal and sophisticated weapon systems were originally designed to
counter the massive military presence of the Warsaw Pact in Europe. With
the dissolution of that threat and, indeed, of the Soviet Union itself, the Army
has little need for more expensive and sophisticated modern weapons. The
Army will soon have considerably more tanks than it needs to equip its
planned forces. Reflecting this excess, the Bush Administration's plan, as
submitted in February 1992, included no funds to buy new tanks after 1992.
At the Congress's direction, the Army will modify older Abrams tanks to a
more capable configuration in fiscal year 1993. The future of the tank
program after 1993, however, is undecided.
Although the lack of need for an active tank line may be welcome news
in an age of constrained military budgets, it also causes concern about the
possibility that the United States could, in the near future, be unable to
produce tanks. This inability may be particularly troubling in a world that
recently witnessed a major war in the Persian Gulf that featured the use of
tanks and other armored vehicles. Questions have been raised regarding the
continued ability of the United States to produce modern tanks in response
to changing world conditions, should an unforeseen threat emerge. In
response to these concerns, members of industry and the Congress have
suggested options for maintaining all or part of the tank industrial base.
This paper describes the U.S. industrial base for producing tanks,
outlines the tank requirements of the U.S. Army, and then presents and
examines options for preserving the tank industrial base in either an active or
a stored mode.
The Tank Industrial Base
The tank industrial base in the United States is an extensive network of

facilities that contribute to the manufacture of the finished tank. The current
base includes major corporations on contract to the government, known as
prime contractors; government facilities, including Army arsenals and
Department of Energy (DOE) plants; and companies, known as subcontrac-
tors, that have been engaged to supply parts and materials to the prime
contractors. Table 1 lists the major facilities that the Army includes in its
definition of the tank industrial base. Since 1987, the sole product of the U.S.
tank base has been the Abrams tank—the Army's newest and most sophisticat-
ed tank.
TABLE 1. OVERVIEW OF THE U.S. TANK INDUSTRIAL BASE
Company or Facility Location Product
Contractors to the U.S. Government
Private Corporations
General Dynamics, Land

Systems Division (GDLS)
Lima Army Tank Plant* Lima, Ohio Tank, hull, turret
Scranton Plant Scranton, Pennsylvania Castings
Sterling Plant Sterling Hgts., Michigan Electric components
Central Office Complex Sterling Hgts., Michigan Management
Detroit Arsenal Tank Plant3 Warren, Michigan Components
Textron Lycominga Stratford, Connecticut Engine
Allison Transmission Division Indianapolis, Indiana Transmission
Hughes Aircraft Corporation Culver City, California Laser range finder,

thermal sight
Computing Devices of Canada Nepean, Ontario Ballistics computer
FMC Corporation Alabama Track
Stanley Illinois Final drive
Urdan Israel Road wheels
Government Facilities
Department of Energy Idaho Falls, Idaho Special armor
Rock Island Arsenal Rock Island, Illinois Gun mount
Watervliet Arsenal Watervliet, New York Cannon
Subcontractors
Cadillac Gage Detroit, Michigan Stabilizationsystem
Rockwell International Kansas Race ring support
Garret AiResearch Torrance, California NBC protection

system

NOTE: NBC = nuclear, biological, chemical,
a. Government-owned, contractor-operated facility.
ALTERNATIVES FOR THE U.S. TANK INDUSTRIAL BASE 10
The major contractor for the Abrams tank is General Dynamics, Land
Systems Division (GDLS). As operator of the government-owned tank plants
at Lima, Ohio, and Warren, Michigan, it is responsible for delivering finished
tanks to the Department of the Army.1 In addition, several other firms have
contracts with the Army to produce major components such as engines or
transmissions, which the Army then turns over to GDLS for incorporation into
the finished tank. Some producers in this category are private companies such
as Textron Lycoming, which supplies the tank engines. Others are Army
installations such as Rock Island Arsenal, which provides the gun mounts, and
Watervliet Arsenal, which makes the cannons. The facility that manufactures
armor inserts out of depleted uranium is operated by DOE. Most of these
facilities also produce other items or perform other services. A few, the DOE
armor facility being one example, are supported solely by work related to tank
production.
A final part of the industrial base is composed of subcontractors who

work under contract to firms with contracts with the Army. This group forms
the most numerous portion of the industrial base, although it may not employ
the most people or do the largest volume of business. It comprises about
18,000 firms, large and small, that provide all kinds of components that go
into the making of a tank, including ball bearings and electronic components
as well as large cast pieces. In general, these companies are not supported
solely by producing parts for tanks. Some units within these companies,
however, may specialize in producing tank parts, and those units might close
if tank production ceases.
The industrial base that supports tank production is extensive and

widespread, involving many facilities located in many parts of the country.
Major facilities in the tank industrial base are spread over 24 states and
Canada (see Figure 1). Facilities located in Michigan, Connecticut, Ohio,
Indiana, and California receive the largest shares of production funding, but
other states also derive significant revenue from tank work done within their
borders (see Table 2).
The Army's Tank Requirements
To fully support its fighting forces, the Army needs many tanks in addition to
those that equip its armored units. The Army must provide tanks for the
Armor School to train tank crews, it must have sufficient spare tanks on hand
to replace those that need repair, and it must maintain reserve stocks to be
1. The tank plant at Warren, Michigan, is also known as the Detroit Arsenal Tank Plant.
FIGURE 1. DISTRIBUTION OF M1A1 PRODUCTION FUNDS
Funds Spent per Tank

D 0-$2,000
H $2,000-$30,000
• $30,000-$300,000
I More than $300, OCX)
SOURCE: Congressional Budget Office based on U.S. Anny, Tank Industrial Base Briefing* (February 20,
1992).
TABLE 2. FUNDS SPENT PER Ml A 1 TANK IN VARIOUS STATES
Funds Soent oer M1A1 Tank

Thousands As a Percent-
State or Region of Dollars age of Total
Michigan 561 18
Connecticut 550 18
Ohio 419 14
Indiana 392 13
California 314 10
New York 142 5
Pennsylvania 128 4
Illinois 93 3
South Carolina 83 3
New Jersey 58 2
Alabama 57 2
Canada 50 2
Massachusetts 45 1
Florida 34 1
North Carolina 23 a
Maine 16 a
Texas 16 a
Virginia 12 a
Oregon 12 a
Missouri 8 a
Iowa 4 a
Minnesota 4 a
Arizona 4 a
New Hampshire 2 a
Wisconsin 2 a
Remaining States _68 _2
Total 3,097 100
SOURCE: Congressional Budget Office based on U.S. Army, Tank Industrial Base Briefing" (February 20,
1992).
a. Less than 1 percent.

used in the event of war. Finally, the Army maintains extra sets of tanks
overseas, primarily in Europe, that would be available for immediate use by
combat units stationed in the United States. These additional requirements
approximately double the number of tanks that the Army needs solely to
equip its armored units in order to support its forces fully.2
Current Inventory. The U.S. Army now has almost 16,000 tanks of all types,
about double the size of its tank fleet in 1974 (see Figure 2). This increase
in the number of tanks was not necessitated by any increase in the size of the
Army's armored forces. Rather, the Army needed the increased inventory
primarily because the United States decided to preposition several divisions'
worth of tanks in Europe, a policy designed to speed the deployment of U.S.
forces in the event of a major war. Additional tanks were also needed to
provide war reserves that were stored in Europe and would be used to replace
tanks lost in combat.
In addition to expanding its tank inventory during the last 20 years, the
Army was purchasing better tanks—specifically, the Abrams, regarded by many
as the best tank in the world.3 The Army's procurement of the Abrams,
which started in 1979, ended in 1992. The Army's current inventory includes
about 7,700 Abrams tanks, the oldest of which is 13 years old. The Army also
has many older and less capable tanks, including about 7,300 M60 tanks that
range in age from 10 to 27 years old, and slightly less than 900 M48s that are
over 30 years old and obsolete (see Table 3).
Smaller Inventory Needed for 1995 and Thereafter. The Army estimates that
it will have 7,780 Abrams tanks in its inventory in 1995. The Congressional
Budget Office calculates that this number of tanks will suffice to equip the
force proposed by the Bush Administration for 1995. Reflecting changed
circumstances in the world, the Bush Administration proposed reducing the
size of the Army so that by 1995 it would be about 30 percent smaller than
it was in 1990. As a consequence, the number of tanks needed to equip the
Army would also be reduced. CBO's analysis assumed current levels for
equipping forces, and proportional requirements—based on current levels—for
tanks needed for training, spares, and war reserves. Based on the Bush
Administration's plans, CBO also assumed the need for 2% divisions' worth of
prepositioned stocks. Using these assumptions, CBO calculated that the
Army's need for tanks will decrease from approximately 11,100 to support
2. This estimate does not include tanks needed by the Marine Corps, which are not addressed in this paper.
3. In his testimony before the Senate Committee on Armed Services, General Colin Powell, the Chairman of
the Joint Chiefs of Staff, stated that the Army's M1A1 tank is unmatched in the world.
FIGURE 2. U.S. ARMY TOTAL TANK INVENTORY

20,000 r
1970 1972 1974 1976 1978 1980 1982 1984 1986 1988 1990 1992
Fiscal Year
SOURCE: Congressional Budget Office based on John M. Collins, U.S.-Soviet Military Balance: Concepts and
Capabilities, 1960-1980 (New York: McGraw-Hill Publications, 1980); International Institute for
Strategic Studies, The Military Balance (London: Brassey's, various editions); and Army data.
28 divisions in 1990, to about 9,600 for the current force. By 1995, the Army
would need fewer than 8,000 tanks to support the 20 divisions that it plans to
retain in its force structure, which includes 12 active, six reserve, and two
cadre divisions.
Although the Army has not announced a formal plan regarding

disposition of its excess tanks, a number of actions could be taken. Some
older vehicles, particularly the M48s, could be scrapped or used for target
practice. Of the 4,900 tanks currently stationed in Europe, almost 900 must
be destroyed, transferred, or sold to allies within the North Atlantic Treaty
Organization. Excess tanks stationed in the United States or Korea could be
sold or given to non-European allies, and some of the excess vehicles could
be retained as additional war reserves. Depending on the Army's approach,
the total inventory of tanks in 1995 will probably exceed the minimum number
needed to equip and support the planned forces. Nevertheless, the remaining
analysis in this paper focuses on only those tanks needed to meet U.S. force
requirements in 1995 and beyond.
The fleet of tanks required by 1995 could be composed entirely of

Abrams tanks and be more capable, more sophisticated, and only slightly
older than today's fleet. The fleet will be an average of 7.9 years old,
compared with 7.4 years today (see Table 3). The required inventory in 1995
also will be more capable on a tank-for-tank basis than today's inventory.
TABLE 3. THE U.S. ARMY TANK FLEET
Age (Years) Capability"

Number of Tanks Young- Old- Aver- Aver-
M48 M60 Ml Total est est age Total age
Total Inventory
1993 896 7,311 7,660 15,867 0 32 12.9 71,000 4.4
Required Inventory
1993 0 1,940 7,660 9,600 0 14 7.4 50,800 5.3
1995 0 0 7,780 7,780 2 15 7.9 44,600 5.7
SOURCE: Congressional Budget Office based on Army data and on The Analytic Sciences Corporation, The
TASCFORM Methodology: A Technique for Assessing Comparative Force Modernization, 4th ed., rev.
(Arlington, Va.: TASC, July 1991).
a. Based on TASCFORM scores.

This conclusion is based on a quantitative measure of each tank's capability

developed by The Analytic Sciences Corporation using a method called
TASCFORM.4 This method assigns scores to types of weapons~the M48,
M60, and Ml tanks, for example-based on characteristics such as their firing
range, type of armor, and fire control apparatus. Thus, an Abrams tank,
which has a longer firing range and a more sophisticated fire control system
than the M60 tank, would have a higher score than the M60. Tanks in the
required inventory in 1995 will have an average TASCFORM score of 5.7,
compared with an average score of 5.3 for tanks in today's required inventory
(see Table 3).
Comparison with Tank Fleets from Other Countries. Even though the U.S.
tank inventory may be shrinking during the next three years, it will still be
more capable than the inventories of most other countries (see Figure 3).
This advantage stems in part from the fact that each tank in the U.S.
inventory will be highly capable, but also from the significant changes that are
expected in the tank inventories of other countries. Just as the United States
must reduce its European tank holdings to comply with the Conventional
Forces in Europe (CFE) Treaty, so too must other European countries. Thus,
for example, the size of Ukraine's and Russia's tank holdings in Europe will
decrease significantly by 1995. And Germany will have to dispose of most of
the equipment it inherited from the East German army.
Based on sheer numbers and discounting the obvious advantage in

technology held by the Abrams tank, the U.S. tank fleet should dominate most
others in 1995. The United States' 7,780 tanks would roughly equal China's
inventory and would be two to four times larger than the inventories of
postwar Iraq, Libya, and North Korea.
The total capability of the U.S. required inventory in 1995-based on the

TASCFORM methodology-would be about one-third less than today's total
fleet because of the reduced number of tanks. Nevertheless, except for the
Russian fleet, the U.S. fleet would still be preeminent in the world. The U.S.
fleet would be five times more capable than China's and up to 12 times more
capable than the tank forces of postwar Iraq, Libya, or North Korea.
Russia will be the only country to possess a tank fleet that will be larger
and more capable than the U.S. Abrams fleet. In 1995, Russia could have
35,000 tanks, whose combined capability would be twice that of the U.S. fleet.
This fleet will be dispersed over the entire Russian expanse, however, which
stretches more than 5,000 miles from the Baltic countries to Siberia.
4. The Analytic Sciences Corporation, The TASCFORM Methodology: A Technique for Assessing Comparative
Force Modernization, 4th ed., rev. (Arlington, Va.: TASC, July 1991).
FIGURE 3. COMPARISON OF VARIOUS TANK FLEETS
ABSOLUTE SIZE OF TANK FLEET
CAPABILITY OF TANK FLEET '
SOURCE: Congressional Budget Office based on data from the International Institute for Strategic Studies, The
Military Balance, 1992-1993 (London: Brassey's, Autumn 1992); The Analytic Sciences Corporation,
The TASCFORM Methodology: A Technique for Assessing Comparative Force Modernization, 4th ed.,
rev. (Arlington, Va.: TASC, July 1991); and Army data.

Furthermore, the country is divided geographically by the Ural Mountains,

with the region east of the Urals considered Asia and that to the west
considered Europe and subject to constraints of the CFE treaty. Transporting
tanks from one end of the country to the other would require great effort, and
forces stationed in Asia have generally been considered unavailable for use
in conflicts in Europe and vice versa.
When considered separately, the two Russian fleets appear less

formidable. Indeed, by 1995, the U.S. Abrams fleet will exceed the European
Russian fleet by more than 20 percent in sheer numbers, and by even more--
about 60 percent~in capability. Compared with the Asian fleet, which is
composed of older and less capable tanks, the U.S. Abrams fleet will be
outnumbered by a factor of almost four but will be at less than a 40 percent
disadvantage in capability.
The U.S. Army therefore does not have an immediate need for new
tanks. It can equip and support the planned forces with Abrams tanks that
have already been purchased. Furthermore, the Abrams fleet is composed of
new tanks; the oldest will be just 30 years old in 2010. Because the Army has
retained tanks in its inventory for 30 years or longer, production to meet
inventory requirements can, in theory, be delayed until 2010 or later. Finally,
the superiority of these tanks was well established by their performance in the
Persian Gulf War.
Future Trends and Their Potential Impact on the U.S. Tank Fleet. Two types
of threat could emerge during the next decade that could compel the Army
to buy new tanks before 2010, although neither seems probable. The first
would be the reemergence of a military entity capable of fielding a large
armored force similar to that of the former Soviet Union. The second would
be the fielding, in significant numbers, of a tank that would be technologically
superior to the Abrams tank. Either occurrence could force the United States
to build and field more sophisticated tanks to counter those of a potential
adversary.
The emergence of a potential foe with a large and threatening armored

force might require the United States to buy more tanks before 2010, but that
event seems unlikely. The former Soviet Union and its allies in the Warsaw
Pact represented such a foe, but few countries in the world today fill this bill.
The merger of several republics of the former Soviet Union into a belligerent
entity could cause concern, but such a merger seems highly unlikely in view
of the contentious relations among the republics and former members of the
Warsaw Pact. Both Russia and China field large tank forces, but both forces
are widely dispersed and, at least in Asia, are composed of relatively
unsophisticated tanks. In Europe, Russian tank forces are limited by treaty
and must contend with the combined forces of the North Atlantic Treaty
Organization.
Moreover, building many new tanks would not be the only way for the
United States to respond to such a threat. A cheaper and more timely
response would be to retain or reconstitute some of the forces scheduled to
be eliminated in the next few years. Such units could be equipped with tanks
currently in the U.S. inventory, such as the M60A3. Unless they are sold,
given away, or destroyed, several thousand surplus M60A3 tanks, which are
not part of the required inventory, will remain in the U.S. inventory for many
years to come. Though not as capable as the Abrams, these tanks are
nonetheless a match for many tanks currently fielded by the former Soviet
Union and almost all tanks fielded by developing countries.
A more plausible scenario, and therefore one of greater concern, would

be the proliferation of highly sophisticated tanks, capable of challenging the
Abrams, in developing countries. Such a threat would take a while to
materialize, however, because no such tank is currently being sold overseas.
The French Leclerc, the German Leopard 2 Imp, and the British Challenger
2 are on a par with one of the later models of the Abrams tank—the
MlAl~but they have not yet been fielded or produced in large numbers.
When they become available, however, they could, if enough were bought,
pose a serious threat. That would be particularly true in areas of the world
where the United States does not have a continuing forward presence or the
capability to deploy a large armored force.
Implications for the Tank Industrial Base
The question that needs to be answered, then, is will the United States need
more than 7,780 capable tanks in the period before a successor to the Abrams
is produced, probably sometime after 2010? Largely as a result of the
diminished threat from the Warsaw Pact and the Soviet Union, the
overwhelming need that motivated many Army programs for so long no longer
exists. However, inasmuch as no one could have predicted the collapse of the
Warsaw Pact and the Soviet Union two years ago, no one can say with
certainty that no new or rejuvenated threat will emerge in their place in the
next 15 years. Maintaining the ability to produce tanks, should they be
needed, can be viewed as an insurance policy for these uncertain times.
Whether it is worth the cost of the premium in this time of fiscal constraint
and how long it is needed are questions that both the Administration and the
Congress need to address. To help in this debate, CBO examined three
options of various costs for maintaining some parts of the tank industrial base
for the foreseeable future.
OPTIONS FOR MAINTAINING THE TANK INDUSTRIAL BASE
The options for preserving the tank industrial base range from putting it in
mothballs after completing currently planned tank production to continuing
tank production at a low rate. CBO examined both of these options but also
considered an intermediate alternative that would modify tanks in the current
inventory to make them more capable. The options would add similar
capability to the active inventory, but they differ in their cost and in the
amount of insurance they provide in retaining the ability to produce tanks if
a crisis arises.
The options also vary in the impact that they would have on the jobs of
people employed in the tank industrial base. Although this issue is of great
concern to all involved, it was not the focus of this paper and so was not
analyzed.
Alternative I: Lay Away the Tank Industrial Base
Realizing that the Army has no near-term need for new tanks, the United
States drastically reduced its annual purchases of Abrams tanks for the Army
from a total of 481 tanks in 1990 to 18 in 1992. After 1992, the Bush
Administration's budget, as submitted in February 1992, includes no funds to
buy new tanks. Following the pattern of declining purchases, the annual
production of tanks has fallen significantly from the 718 produced in 1990 to
the 484 produced in 1992. Assembly of tanks for the United States will
continue through 1993. After that time, no new tanks will be produced to fill
U.S. orders until the next generation is needed. At the Congress's direction,
the Army will modify 54 older Abrams tanks to a more capable configuration
starting in 1993.
Plans for the tank program during the next five years are uncertain. The
Army has proposed a conversion program in two phases, the first of which
would upgrade 210 tanks and would require funding in 1994 and 1995. The
second phase would upgrade almost 800 additional tanks and would require
funding for several years after 1995. The revised budget submitted by then
Secretary of Defense Dick Cheney in January 1993 fully funded the first
phase, but provided only minimal funds for the second. Furthermore, the
Bush Administration did not set a firm date for when the Army will start
producing its next tank, although the current inventory of Abrams tanks will
begin to reach the end of its estimated useful life in 2010.
In the meantime, production to fill foreign orders will continue. General

Dynamics' Land Systems Division is under contract to supply Egypt with 555
kits for producing Abrams tanks. Saudi Arabia has agreed to buy 465 Abrams
tanks, and Kuwait has recently ordered 236 tanks. These sales could keep
U.S. tank production going through at least 1996. The Bush Administration
also investigated the possibility of additional foreign military sales~a total of
783 tanks-that would maintain production at least through 1999. Potential
additional customers for these tanks are the United Arab Emirates and
Sweden. None of these additional sales is firm, however, and even GDLS is
not confident that they will all come to fruition. Furthermore, the political
cost of adopting the role of arms merchant while calling for limits on
worldwide arms sales would have to be weighed against potential benefits.
Finally, the technological risk of proliferating sophisticated U.S. systems
around the world should also be considered. Maintaining the U.S. tank base
through foreign military sales may not be a prudent long-term solution.
Description. As production of tanks or their major components ceases, one

option would be to put the government-owned production equipment and
facilities into storage until needed again.5 This process, known as layaway,
is a progressive one. For example, when production fell below 50 tanks per
month, it was no longer necessary to continue operating the two tank
assembly plants~the Detroit Arsenal Tank Plant in Warren, Michigan, and a
plant in Lima, Ohio. Consequently, the assembly line in Warren was closed
in September 1991 and is now being laid away. Furthermore, because the
tanks sold to foreign buyers do not include the special armor found on the
latest model of the U.S. tanks, the DOE facility for producing this armor will
close in 1993 or 1994.6 The DOE equipment and facility will then be laid
away until they are needed again.
Should the United States cease producing tanks in 1997, one option
would be to lay away the additional major facilities associated with tank
production that would no longer be needed. The manufacturing equipment
unique to tank production would be stored in place and protected from the
elements, if necessary. Government-owned buildings would be closed and
guarded. Leased facilities would continue to be rented so that they would be
available if needed. In short, the physical portions of the tank production
base would be stored for future use. While preserving the production
5. The Bush Administration proposed this course of action in its February 1992 budget submission. The proposed
budget revisions submitted in January 1993 by then Secretary of Defense Dick Cheney include funds for the
tank conversion program described earlier. The Clinton Administration is currently considering its plans for
the tank program.
6. Without the program to convert 54 tanks in fiscal year 1993, the DOE facility would have closed in December
1992. Because the facility can produce 20 armor sets per month, it could be ready for shutdown sometime in
fiscal year 1993.
facilities, the Army could continue to develop a tank to replace the Abrams
fleet starting in 2010.
Time and Cost to Restart Production. When production resumes and the
industrial base is needed again, time will be needed to take it out of
mothballs and to find and train the people to run it. If the line remains shut
for more than two years, restarting it could take a long time. First of all,
some skilled personnel in jobs that require extensive training would, according
to the Army, have to be retrained to regain their skills.7 Retraining for some
tasks, such as armor welding, could take six months, according to Army
estimates.
Second, the Army claims that, after a hiatus of two or more years, the
DOE facility could take two years to be recertified and start producing special
armor. (Of course, if an urgent need arose for tanks to meet a national
emergency, this recertification requirement could be waived or expedited.)
After certification, producing the first set of armor would take another 18
months, for a total delay of 42 months. According to the Army's analysis, the
special armor is responsible for the longest delays when restarting an inactive
tank line. Thus, all the other components should be available when the armor
is ready.
Finally, another nine months would be needed to assemble all the parts
into the first tank. In all, it could take up to 51 months to rejuvenate the
industrial base and produce the first tank. In a time of crisis, however, the
Army will need tanks in large numbers. Increasing production rates takes
time because more trained workers are needed and shifts are added. In fact,
attaining a production rate of 120 tanks per month could take an additional
21 months. Thus, a total of six years might elapse before tanks would be
produced in large numbers.
A one-time investment will be needed to take the production line out of

mothballs. The one-time costs to restart production and achieve rates of 120
Abrams tanks per month, including training costs for people in highly skilled
jobs, would be about $1.1 billion in 1993 dollars, according to Army
estimates.8
7. Retraining time might be minimized if the government developed and maintained an up-to-date list of the
addresses of personnel with skills critical to the manufacture of tanks. Such a list might be maintained by
offering people with critical skills a modest remuneration in return for keeping the government informed of
their whereabouts. Alternatively, the contractors or the Army could establish schools to maintain the critical
skills of people who would no longer be involved in tank production.
8. The estimate of $1.1 billion includes only training and start-up costs. Production would require additional
funds.
Cost to Lay Away. The Army estimates that laying away and storing the tank
industrial base will cost $340 million in 1993 dollars over the next five years
(see Table 4). That amount includes $200 million to meet the one-time costs
associated with ceasing tank production and laying away the tank facilities
(see Table 5). One-time costs would cover personnel separation, laying away
facilities and equipment, managing the layaway, and environmental cleanup.
Costs associated with management and personnel separation form the largest
component of one-time costs (45 percent); the remaining 55 percent would be
devoted to layaway and cleanup. The Army admits, however, that its analysis
of potential environmental costs is not complete and that those costs could be
much higher.
Laying away the tank production base would also create annual recurring
costs for as long as the lines are in storage. From 1994 through 1998,
recurring costs for security and maintenance of the laid-away facilities
TABLE 4. ANNUAL COSTS THROUGH 1998 FOR VARIOUS ALTERNATIVES

1993 Total
and 1994-
Before 1994 1995 1996 1997 1998 1998
Alternative I:
Lay Away Tank Line 0 50 30 140 80 40 340
Alternative II:
Produce at Low Rate* 0 780 800 760 740 750 3,830
Alternative III:
Convert Older Modelsb
Phase I 680 80 130 0 0 0 210
Phase II a ' c 0 0 150 550 650 650 2,000
NOTE: Costs do not include about $60 million to lay away the assembly line at the Detroit Arsenal Tank Plant.
a. Assumes additional production of 120 tanks per year to fill orders from foreign governments.
b. Includes estimates for research and development, procurement, and operations and maintenance costs to
convert Ml tanks to the M1A2 configuration.
c. According to then Secretary of Defense Dick Cheney's proposed budget revisions submitted in January 1993,
a minimum annual investment of $250 million would sustain the program after 1995, but at 60 conversions per
year rather than the 198 conversions funded in the alternative.
TABLE 5. COSTS FROM 1994 THROUGH 1998 FOR ALTERNATIVE I (LAY AWAY THE
TANK LINE), BY FUNCTION AND FACILITY (In millions of 1993 dollars)
Detroit Lima
Arsenal Army Allison
Tank Tank Trans-
Function Planta Plant GDLSb mission Other0 DOE Total
One-Time Costs
Management and
Personnel Separation 20 40 30 90
Layaway of Facilities
and Equipment 10 30 20 d 10 0 70
Environmental Cleanup 10 J2 _d d _0 JO JO
Subtotal 40 80 50 d 10 20 200
Recurring Costs
Security and
Maintenance 20 10 20 90 140
Total
One-Time and
Recurring Costs 40 100 60 20 10 110 340
SOURCE: Congressional Budget Office based on U.S. Army Tank-Automotive Command, Austere Abrams Closure
Study (August 1991); and personal communication with Tank-Automotive Command, March 1992.
NOTES: GDLS = General Dynamics, Land Systems Division; DOE = Department of Energy.
a. Cost associated with laying away component machining only. Excludes an additional $15 million associated
with laying away the assembly line.
b. Includes Central Office Complex, Scranton Plant, Sterling Plant, subcontractors, and vendors.
c. Includes specific product lines operated by Textron Lycoming at the Stratford Army Engine Plant and other
lines at Hughes Aircraft Company, Computing Devices of Canada, Rock Island Arsenal, Watervliet Arsenal,
Kollmorgen, Plessey, and Singer-Kearfott.
d. Less than $5 million.

would amount to $140 million, or 41 percent of the total five-year costs. After
1998, and until facilities would be needed to start producing parts for the next
generation of Army tanks, annual costs for maintaining and preserving the
base would total about $40 million (in 1993 dollars).
Ceasing tank production in the United States would affect many

facilities throughout the country, but the Army has compiled cost estimates
for only the largest. A breakdown of the layaway costs by facility indicates
that the DOE facility and the Lima tank plant would incur the largest costs
(see Table 5). Closing, maintaining, securing, and cleaning up the DOE
facility could cost $110 million from 1994 through 1998, with the bulk of this
amount devoted to caretaking. Closure and layaway of the Lima tank plant
would cost $100 million during this same period. The assembly line at the
Detroit Arsenal Tank Plant (DATP) at Warren was closed in September 1991
when both it and the assembly facility at Lima were no longer needed to fill
orders. The DATP still continues to manufacture some components. The
Army estimates that $49 million will be needed from 1994 through 1998 to lay
away the entire DATP. Some portion of this amount, however, will be
needed even if tank production does not cease, because the production rate
has already fallen below the level needed to sustain two tank assembly lines.
Indeed, about $3 million will be needed annually to lay away the assembly
line at the DATP as long as production remains below about 600 tanks per
year. Since rates this high are not anticipated, even with the most optimistic
expectations for foreign military sales or under any of the options examined
by CBO, this cost would be incurred no matter which of the alternatives is
adopted; it therefore is not included in the discussion of costs associated with
the options.
The costs for closing and storing the remaining dedicated lines would be
spread over several facilities. The major remaining facilities include the
GDLS facilities (the Scranton castings plant and the plant at Sterling Heights,
Michigan), the transmission line at Allison Transmission Division, and the
engine line at the Stratford Army Engine Plant operated by Textron
Lycoming. Finally, some smaller lines that produce parts that the government
supplies to GDLS for incorporation into the tank, such as laser range finders
from Hughes, would also have to be closed, stored, and maintained.
Alternative II: Maintain Production, but at a Low Rate
One way of maintaining the tank industrial base is to continue to produce

tanks but at a lower rate. To maintain all parts of the base, including the
DOE facility, some tanks will have to be produced for the U.S. military,
because tanks sold to foreign customers do not contain the special armor that
the DOE facility produces.
Description. In this option, the Army would buy 120 of the latest model
Abrams tank (the M1A2) per year for the U.S. military, starting in 1994.
Production at the rate envisioned in this option will not guarantee that all
parties currently producing tanks or their parts will continue to do so. For
some of the 18,000 vendors who now supply or have supplied components for
tanks, the low volume of business that would result from this option may not
be sufficiently profitable to keep them in the tank business, particularly
because many tank parts must meet exacting government standards.9 In fact,
because the rate of tank production has decreased over the past two years,
several vendors have already ceased their involvement in the tank program,
and GDLS expects that several more will follow. Nevertheless, producing
tanks at a low rate will sustain all major parts of the industrial base.
Costs. This alternative would be quite expensive to carry out. Low-rate

production is inherently inefficient because facilities that were designed to
produce items in large numbers—the active tank base has the capacity to
produce 75 tanks per month-are not fully used at lower rates. The few tanks
being produced, however, must bear the overhead costs for the unused
capacity. The cost per tank therefore goes up, and relatively high prices are
paid for just a few tanks. Based on Army estimates, the cost to produce one
tank increases by almost 70 percent as the production rate decreases from 360
tanks per year to 120 tanks per year. The cost over the next five years to buy
600 tanks would be about $3.8 billion. (These costs are based on the
assumption that an additional 120 tanks per year will be produced for foreign
buyers. Without the foreign sales, costs for U.S. tanks will increase above
those cited here.) After 1998, annual costs would be about $740 million.
Should new tanks ever be needed in large numbers, the time and cost to
reach a surge rate of 120 tanks a month would be 47 months and about $600
million.
Alternative III: Convert Older Abrams Tanks to Newer Models
This alternative represents a compromise between laying away the

components of the tank base and continuing production. It would maintain
most of the industrial base by converting older, less capable versions of the
Abrams tank to the newest and most capable model. The Congress has
supported this approach for several years and has provided funds to carry it
9. The government-owned tank assembly facility at Lima, Ohio, has the capacity to produce 75 tanks a month
(900 tanks a year)—much more than the number of tanks envisioned in this alternative.
out. Neither the Bush Administration nor the Army included such a program
in the budget submitted in February 1992 or advocated conversion during
subsequent hearings on the fiscal year 1993 budget request. In the fall of
1992, however, the Army began to evaluate a program that would upgrade
1,002 Abrams tanks and has since submitted its proposed program to the
Office of the Secretary of Defense for approval. The Bush Administration
had approved only the first phase of the program by the end of its tenure, and
the Clinton Administration has not yet made its position on this program
known. This alternative would carry out the Army's plan.
A conversion program might spur prospects for foreign military sales,

which would ensure support for those parts of the base not needed for a
conversion program. It would also provide a hedge against unforeseen
circumstances that would require the Army to modernize its tank inventory.
Finally, this alternative, though costing more than laying away the industrial
base, would cost less than keeping the production lines open; if pursued for
a sufficient length of time, it would bring almost all tanks in the active
inventory up to the same high level of capability.
Description. Four models of the Abrams tank have been produced since
1979. In chronological order and order of increasing capability, they are the
Ml, the IPM1, the M1A1, and the M1A2. The first two models have a
105mm cannon, and the last two a 120mm cannon with longer range. Other
improvements have been added over the years, including armor made with
depleted uranium (see Table 6). All M1A2 tanks and about half of the
MlAls have this special armor.
This option would include a program to convert the oldest Abrams

tanks—the Mls--to the M1A2 configuration. Such a modification would entail
replacing the entire turret, including the gun; installing new electro-optics;
modifying the hull so that it can accept the active nuclear, biological, and
chemical (NBC) protective system; and replacing the armor packages and the
entire electrical system.
Such a conversion process would involve most components of the tank

industrial base. According to analysis performed by GDLS, this modification
process would require 88 percent of the critical skills involved in producing
tanks and would maintain 100 percent of the unique manufacturing elements
now employed in tank production.10 The engine and transmission in the
10. GDLS's analysis identified the processes that are unique to tank production and not easily transferred to
civilian application. Those processes would not include engine and transmission work that, though different
for use in tanks, is not sufficiently different to make it unique.
TABLE 6. CHARACTERISTICS OF THE ABRAMS TANK INVENTORY
Characteristic Ml IPM1 M1A1 M1A2
Years Produced 1979-1985 1984-1986 1985-1993 1992-1993
Number in Army
Inventory in 1995 2,374 894 4,430 62
Weight (Tons) 60 61 67 69
Caliber of Main Gun

(Millimeters) 105 105 120 120
Armor Standard Standard Standard/ Special

Special3
Stored Rounds
(Main gun) 55 55 40 40
Maximum Road Range

(Miles) 280 310 275 265
Improvements over n.a. Improved Further Commander's

Previous Model Suspension Improved Independent
Suspension Thermal
Improved Viewer
Armor Active NBC
Protective Digital
System Electronics
SINCGARS
Radiob
Improvement in
Capability over
Ml Model0 (Percent) n.a. 4 7d 22
SOURCE Congressional Budget Office based on Christopher F. Foss, ed., Jane's Armour and Artillery Systems
(Surrey, England: Jane's Information Group, various editions); Association of the U.S. Army, Army,
1991 Greenback (October 1991); data from General Dynamics, Land Systems Division; and The
Analytic Sciences Corporation, The TASCFORM Methodology: A Technique for Assessing
Comparative Force Modernization, 4th ed., rev. (Arlington, Va.: TASC, July 1991).
NOTES: NBC = nuclear, biological, chemical; SINCGARS = Single Channel Ground and Airborne Radio
System; n.a. = not applicable.
a. Standard armor on 2,266 MlAls and special armor on the rest.

b. If available.
c. Based on TASCFORM scores.
d. With special armor.
Abrams is common to all models, however, so no new engines or transmis-

sions would be needed to upgrade Mis to the M1A2 configuration. Some
analysts have argued that if the United States employs a significant number
of the M1A2 model in its forces, other countries will be interested in buying
new M1A2 tanks. Indeed, Saudi Arabia and Kuwait have already ordered a
total of 701 MlA2s. Thus, foreign sales of new MlA2s might keep Allison
Transmission and Textron Lycoming in the tank transmission and engine
business, at least for a while.
The Army, in response to Congressional prodding, designed a specific

program in the fall of 1992 that would convert 1,002 Abrams tanks to the
M1A2 configuration starting in 1993 and ending in 2000. The goal of the
Army's proposed program is to provide the best tanks for potential regional
contingencies. To that end, the converted tanks would be used to equip part
or all of the Army's contingency force, which includes the units that would be
used first in a regional war such as Operation Desert Storm.
The Army has broken its proposed program into two phases. The first
phase would provide enough M1A2 tanks to equip and support one division
within the contingency force and would be funded from 1993 through 1995
(see Table 7). The Army needs almost 300 tanks to equip one mechanized
division and provide tanks for training. The Army has already bought 81 new
M1A2 tanks and plans to convert 210 tanks to the M1A2 configuration in the
first phase of its conversion program. The Army's second phase would
convert an additional 792 tanks and would require funds beginning in 1995.
The length of the second phase depends on the rate at which tanks would be
upgraded. The earliest completion date likely is 2000, which requires
conversion of 198 tanks a year from 1996 through 1999. A less ambitious
program would take longer. Upon completion of both phases, the Army
would have enough M1A2 tanks to equip its planned contingency force (which
includes five divisions plus one brigade afloat) and to provide extra tanks for
training.
Time to Restart Production. The time required to start producing new tanks
from a line that had been engaged in upgrading Mis to MlA2s would be
between the requirements of the previous two alternatives. All portions of the
industrial base would be active except those that produce engines and
transmissions. Of these two, engine production would take longer to restart,
with a delay of 20 months before the first engine could be delivered. Delivery
of the first tank could occur nine months later, for a total delay of 29 months
to produce the first tank. About one year later, production rates of 60 tanks
a month could be expected, and surge rates of 120 tanks could be achieved
about two years after the first tank rolled off the production line. All in all,
it could take 56 months to reach surge-rate production from a line that had
been engaged in upgrading Mis to MlA2s.
Costs. Because the conversion of Ml tanks to the M1A2 model requires such
extensive reworking of each tank, this upgrade would be expensive. In fact,
the total cost of the 210 upgrades in the Army's first phase is $890 million
(see Table 7). If the Army pursues the second phase at a rate of 198
upgrades per year, then the cost for the conversion program from 1994
through 1998 would be about $2.2 billion, with annual costs as high as $650
million (see Tables 4 and 7). Additional annual costs after 1998, if the
program were continued indefinitely at a rate of 198 tanks a year, would be
about $650 million. (As with the previous alternative, these costs are based
on assumed sales to foreign governments of 120 tanks per year. Without
these sales, costs could increase.)
TABLE 7. COSTS THROUGH 1998 FOR ALTERNATIVE III

(CONVERT Mis TO MlA2s)
1993
and
Before 1994 1995 1996 1997 1998 Total
Cost3 (Millions of 1993 dollars)
Phase I 680 80 130 0 0 0 890

Phase Hb 0 0 150 550 650 650 2,000
Number of Tanks Modified
Phase I 54 101 55 0 0 0 210

Phase IIb 0 0 0 198 198 198 594
SOURCE: Congressional Budget Office based on Department of the Army, "Abrams Upgrade Program"
(September 10, 1992); Congressional action on the 1993 budget request; and proposed budget
revisions submitted by then Secretary of Defense Dick Cheney in January 1993.
a. Does not include about $60 million to lay away the assembly line at the Detroit Arsenal Tank Plant. Includes
costs of research and development, procurement, and operations and maintenance to convert Ml tanks to the
M1A2 configuration. The costs of the tanks that will be converted in 1994 and 1995 and perhaps 1996 will
include some of the funds provided in Congressional appropriations from 1993 and prior years. Also, the
Army assumed that it can sell 120 new tanks to foreign customers every year and that producing these tanks
will lower U.S. costs. CBO estimates that if no additional new production from foreign sales occurs after 1995,
total costs could increase by 10 percent to 25 percent.
b. According to then Secretary of Defense Dick Cheney's proposed budget revisions submitted in January 1993,
the conversion program could be maintained with minimum funding of $250 million per year—enough to
convert 60 tanks, assuming that foreign military sales account for 120 new tanks produced annually.
The costs of the conversion program could be lower if the Army pursues
the second phase of the program at a slower pace. In then Secretary of
Defense Dick Cheney's proposed budget revisions, submitted in January 1993,
only $250 million a year is set aside for tank conversion beyond 1995.
Coupled with sales of 120 tanks per year to foreign governments, this funding
should sustain the tank line. The pace of conversion might, however, be quite
slow. The Army indicates informally that it has not yet determined how many
tanks it could convert with $250 million while it also pays the overhead and
other costs necessary to sustain the tank production line.
It seems likely, therefore, that after 1995 the annual costs of a

conversion program will be at least $250 million and could be as high as $650
million. The cost associated with gearing up to a surge rate of production of
new tanks during a national crisis would be about $530 million.
During the next few years, some of the costs of the conversion program
may be financed using either funds from past appropriations or proceeds from
sales of military weapons to foreign countries. For example, the Army
proposes using almost $200 million from the sale of surplus tanks and other
equipment to help pay for the first phase of its conversion program. This
financing approach would reduce the additional funds that the Congress would
have to appropriate to carry out the conversion program. If the conversion
program were not carried out, however, the funds used to finance it would be
available to pay for other needs. The method used to finance the program
therefore does not alter the true costs of converting tanks, which are the focus
of this paper.
COMPARING THE ALTERNATIVES
The three alternatives vary little in the amount of capability they would add
to the Army's tank force. They differ significantly, however, in their costs and
in the amount of insurance each would provide against uncertainty about
future threats to U.S. security.
Arguments in Favor of Layaway
Laying away the tank line offers some significant advantages, especially with
regard to cost.
Least Expensive. Closing and laying away the tank line would require $340
million in funding from 1994 through 1998, much less than the requirements
for maintaining low-rate production ($3.8 billion) or a conversion program (as
much as $2.2 billion) (see Table 8). Annual costs after 1998 would also be
much lower; those associated with layaway would amount to $40 million a
year compared with more than $700 million a year for low-rate production
and from $250 million to $650 million for conversion.
Moreover, the differences in costs could be even greater. In particular,

the costs of low-rate production or conversion may grow beyond current
estimates. Both of these options would require maintaining a company in
business along with its subcontractors and vendors. Because the company
would have to meet overhead costs, estimated costs could grow-perhaps
sharply~if the company's other business (such as foreign military sales)
declines. In contrast, the costs of layaway are less likely to increase. Layaway
involves mothballing equipment and continuing leases and security protection.
Therefore, estimates of these costs should not be subject to much
unanticipated growth.
TABLE 8. COSTS AND RESTART TIMES OF VARIOUS ALTERNATIVES
Cost to
Cost, Time to Reach Surge
1994-1998* Restart Capacity
(Millions of (Months) (Millions of
1993 dollars) First Tank Surgefc 1993 dollars)
Alternative I:
Lay Away Tank Line 340 51 72 1,110
Alternative II:
Produce at Low Rate 3,830C 22 47 600
Alternative III:
Convert Older Models 2,210 c>d 29 56 530
a. Does not include about $15 million to lay away the assembly line at the Detroit Arsenal Tank Plant.
b. Surge production at 120 tanks per month.
c. Assumes additional production of 120 tanks per year to fill orders from foreign governments.
d. The proposed budget revisions submitted by then Secretary of Defense Dick Cheney in January 1993 include
an estimate of $720 million as a minimum sustaining level.
Improvements in Capability May Not Be Needed. Laying away the tank line
would, of course, preclude making any improvements to the capability of the
Army's tank fleet for the foreseeable future. Low-rate production or
conversion would result in increased capability, but the improvements would
be small and may not be needed to meet future threats to U.S. security.
The greatest improvement in the capability of the tank fleet required to

equip and support Army forces would result from carrying out the conversion
program that the Army proposed in the fall of 1992 (see Table 9). Replacing
1,002 Ml tanks with the M1A2 model would result in a 3 percent increase in
capability by the year 2000. Low-rate production, because it could not deliver
as many new M1A2 tanks by the year 2000 as could the modification program,
would result in a somewhat smaller increase in capability (about 2 percent).
But does the Army really need to improve the overall capability of its
tank fleet? Many defense analysts contend that the United States already has
enough tanks to equip its planned forces and that the MlAl is the most
capable tank in the world. There is some basis for this assertion. First, as
mentioned above, U.S. tank forces enjoy a substantial advantage when
compared with tank forces of almost any other country (see Figure 3).
Moreover, by its performance in the Persian Gulf War, the Abrams tank
TABLE 9. EFFECTS OF VARIOUS ALTERNATIVES ON THE ACTIVE

INVENTORY AND CAPABILITY OF TANKS IN 2000
Composition Total
Ml IPM1 MlAl M1A2 Capability a
Alternative I:
Lay Away Tank Line 2,374 894 4,453 62 42,900
Alternative II:
Produce at Low Rate 1,774 894 4,453 662 43,700
Alternative III:
Convert Older Models 1,372 894 4,453 1,064 44,200
SOURCES: Congressional Budget Office based on Army data and on The Analytic Sciences Corporation, The
TASCFORM Methodology: A Technique for Assessing Comparative Force Modernization, 4th ed., rev.
(Arlington, Va.: TASC, July 1991).

demonstrated that it is more capable than its widely deployed counterpart

from the former Soviet Union, the T-72 tank. By all accounts, the Abrams
tank outranged and outmaneuvered the T-72, which is not equipped to fight
effectively at night.
Restart Times May Be Acceptable. If the tank line is laid away, the time and
investment required to restart production and reach production at surge rates
of 120 new tanks a month would be substantial—a full six years and more than
$1 billion (see Table 8). The delays and initial investments associated with
the other two alternatives would be smaller~56 months and $530 million to
achieve surge-rate production from a line upgrading tanks, and only 47
months and $600 million from low-rate production.
The differences in restart times may not, however, be critically

important. Even with low-rate production—the alternative resulting in the
shortest delay-the Army would not be able to add significantly to the tank
force within the likely duration of a brief conflict that developed with little
warning, such as Operation Desert Storm. If a major new threat emerges,
such as a resurgent threat from the republics of the former Soviet Union, it
may well appear with enough warning so that the differences in restart times
are not of great importance.
Having assessed these various arguments, the Deputy Secretary of

Defense in the Bush Administration concluded in April 1992 that the tank
line should be laid away. Donald Atwood stated that the United States now
has enough capable tanks to meet any contingency and that there is enough
time to reconstitute the tank industrial base if a global threat emerges.11
Arguments in Favor of Low-Rate Production or Conversion
Although laying away the tank line has advantages, particularly relating to
cost, low-rate production or conversion provides a hedge against uncertainty
about future threats and other factors. Of the two approaches, conversion
may be preferable because it would maintain much of the industrial base but
would cost at least 40 percent less during the 1994-1998 period. The
remainder of this discussion therefore focuses on a program to convert Ml
tanks to the M1A2 configuration.
Insures Against Uncertainty About Future Threats. In this period of

international instability, when European and Asian peoples long suppressed
by authoritarian regimes are expressing and acting upon nationalist feelings,
11. Testimony of Donald Atwood before the House Committee on Armed Services, April 28, 1992.
no one can be sure if or when a serious new threat to Western security might
emerge. If a new threat emerges, a relatively rapid return to production of
high-quality tanks could be important, especially if considerable time elapses
before the United States recognizes the emergence of and acts to counter a
new military power. A conversion program would require less time to achieve
production of new tanks than would a line that had been laid away.
Even if the emergence of a major new threat is judged too unlikely to

justify long-term maintenance of the tank base, the situation in the republics
of the former Soviet Union could argue for maintaining that base for at least
a few years. Russia alone has about 11,000 tanks in Europe, some of which
are reasonably capable. If a military dictator regained power, Russia could
conceivably reemerge as a military threat to the West within a few years.
Perhaps the most troublesome scenario involves the arming of

developing nations with sophisticated and modern tanks from the global arms
market. As U.S. allies reduce the size of their ground forces, European arms
makers will look outside Europe for potential buyers. Some of the European
manufacturers will eventually be able to produce significant numbers of tanks
sophisticated enough to challenge the M1A1 on the battleground. Thus,
nations in regional hot spots might be willing and able to amass large
inventories of these technologically advanced vehicles.
Compared with these developing nations, the United States will almost
certainly maintain a large numerical advantage in terms of total tank
inventories. The United States might not, however, be able to bring its entire
tank inventory to bear. In regions of the world where the United States does
not maintain a significant forward presence, small units with limited numbers
of U.S. tanks might have to oppose an adversary who threatens U.S. security
interests. Alternatively, the United States might need to split its forces in
order to deal with two or more contingencies simultaneously. In such a
scenario, having U.S. tanks that are superior on a tank-for-tank basis may be
important. U.S. actions in Operation Desert Storm-in which the U.S. military
tried to ensure that its forces used the latest model M1A1, equipped with
special armor-underscore the importance of such superiority.
Converting older Ml tanks to the M1A2 configuration would increase

the chances that U.S. forces would have an advantage on a tank-for-tank basis
in any limited conflict. Indeed, the Army plans to use MlA2s from the
conversion program to equip its contingency force, which includes the units
most likely to fight in a regional war. A conversion program would also
maintain most of the industrial base, thus insuring that the United States will
be able to continue to upgrade its tanks and always equip its forces with the
world's most capable tank.
Insures Against Uncertainty About Time to Restart. Maintaining a tank

conversion program might also hedge against uncertainty regarding the time
required to resume production of quality tanks. The United States does not
have any recent experience in restarting tank production and so has no way
to predict reliably how long it would take to retrain workers and, perhaps of
even greater concern, to reestablish the ability to produce tanks with few or
no defects.
If more vendors than expected go out of business, finding suppliers for

all the parts needed for tanks could take even longer. Moreover, after being
reactivated, a line that has not been producing tanks for a number of years
could take some time to produce quality tanks in large numbers. A couple
of years might be needed to work out the production bugs and to establish a
vendor base capable of providing components of high quality so that tanks
without defects could be produced. Recent production experience with the
Abrams reinforces this concern. According to GDLS, the defect rate for
Abrams tanks coming off the production line has decreased 82 percent during
the past six years. Many problems are associated with starting up new
production lines, to which this would be akin. A high rate of defects in tanks
produced for use in a crisis is obviously an outcome to be avoided.
The uncertainty concerning the ability to resurrect the tank base will
certainly increase the longer the base is shut down. Inasmuch as the United
States will probably not be building new tanks before 2010, the hiatus between
shutdown and reopening may be very long. If restart takes longer than the
Army estimates, then the risks of failing to meet a future threat are greater
than the Administration currently acknowledges.
A conversion program would ensure the availability of quality

components as long as it continued. Even if it continued for only a few years,
a conversion program would also provide additional time to study the issues
related to closing down the production line.
CONCLUSION
In its defense authorization act for 1993, the Congress approved the use of
$225 million-funds that had been appropriated for 1992 but not spent-for
converting older Abrams Ml tanks to the newer A2 version. Thus, at least
through 1993, the Congress opted to insure against possible future threats to
U.S. security by sustaining most of the tank industrial base.
This decision, however, raises the question of how long the United
States should continue to buy insurance. Should it continue until it needs to
produce a new tank, which might not be until 2010 or even later? Or will this
insurance be needed for only a few years until the international climate,
particularly the situation in the former Soviet republics, becomes clearer?
If costs are the key criterion, the tank line should be allowed to close
as soon as all foreign orders are filled. The Army's planned conversion
program would require more than four times as much funding over the next
two years as layaway. Subsequently, depending on the pace of conversion,
upgrading tanks could cost from as little as two to more than 10 times as
much as layaway costs in a given year. Indeed, annual costs for conversion
after 1995 could be as high as $650 million. The Army estimates that the
one-time cost of restarting and reaching surge-rate production from a line that
has been laid away is only slightly more than $1 billion. Thus, if conversion
is pursued at a high pace after 1995, the United States would spend in two
years as much to convert tanks as it would to restart a line that had been laid
away.
In contrast, completing the first phase of the Army's conversion program

would cost a total of only $140 million more in 1994 and 1995 than layaway,
and would result in a useful number of MlA2s. By completing the first phase,
one division of the contingency force could be equipped and supported with
the improved M1A2 by 1996. Ending the conversion program short of
providing enough M1A2 tanks for an entire division would leave the Army in
a dilemma concerning how to employ the MlA2s in its tank fleet. To ease
maintenance and operations in time of war, combat units are equipped
whenever possible with only a single version of each type of weapon. It would
be difficult, therefore, for the Army to use less than a full division's
complement of MlA2s.
After 1995, annual costs for the conversion program could rise to
between $250 million and $650 million. Such costs would represent a small
fraction of the Army's total budget. Sustaining the capacity to produce tanks
through conversion or low-rate production provides some insurance against
an unforeseen event—the emergence of a major new threat to U.S.
security~and other risks associated with closing the tank line. Continuing this
program may therefore be reasonable, particularly until uncertainty about the
situation in Eastern Europe and other parts of the world diminishes.
In the years to come, however, the risks incurred by ending support for
the tank base should continue to be weighed against the costs. This judgment
may be particularly important if external factors, such as a decline in the
volume of foreign sales of tanks, cause the price of the conversion program
to increase.

93doc04 0

Uploaded by

Copyright:

Available Formats

93doc04 0

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

93doc04 0

Uploaded by

Copyright:

Available Formats

CBO

CONGRESSIONAL BUDGET OFFICE

CONGRESSIONAL BUDGET OFFICE

All costs are expressed in 1993 dollars of budget authority.

All years, unless otherwise noted, are fiscal years.

Numbers in tables may not add to totals because of rounding.

Frances M. Lussier prepared this paper under the general supervision

INTRODUCTION AND BACKGROUND 8

The Tank Industrial Base 8

OPTIONS FOR MAINTAINING

Alternative I: Lay Away the Tank Industrial Base 20

COMPARING THE ALTERNATIVES 31

Arguments in Favor of Layaway 31

S-l. Costs of Alternatives for Preserving the

S-2. Annual Costs Through 1998 for

1. Overview of the U.S. Tank Industrial Base 9

2. Funds Spent per M1A1 Tank in Various States 12

3. The U.S. Army Tank Fleet 15

4. Annual Costs Through 1998 for Various

5. Costs from 1994 Through 1998 for Alternative

6. Characteristics of the Abrams Tank Inventory 28

7. Costs Through 1998 for Alternative III

8. Costs and Restart Times of Various Alternatives 32

9. Effects of Various Alternatives on the Active

1. Distribution of M1A1 Production Funds 11

2. U.S. Army Total Tank Inventory 14

3. Comparison of Various Tank Fleets 17

The tank industrial base in this country involves facilities operated by

The Congressional Budget Office (CBO) examined three ways of

Alternative I: Lay Away the Tank Industrial Base

be completed in 1997 with the shutdown of the assembly line. Additional

This plan, espoused by the Bush Administration as recently as April

This required inventory of roughly 8,000 tanks would be more capable

Although the time required to produce significant numbers of tanks

Of more concern, perhaps, is the immediate loss of the ability to build

SUMMARY TABLE 1. COSTS OF ALTERNATIVES FOR

SOURCE: Congressional Budget Office based on Army data.

b. Assuming the program continues indefinitely.

Alternative II: Maintain Production, but at a Low Rate

Alternative III: Convert Older Abrams Tanks to Newer Models

SUMMARY TABLE 2. ANNUAL COSTS THROUGH 1998

a. Assumes annual sales to foreign militaries of 120 tanks.

However it is structured, a conversion program would avoid producing

Should the United States Maintain Its Tank Industrial Base?

1995, conversion would cost more than twice as much as layaway.

In conclusion, reducing risks by maintaining most of the tank base after

During the 1980s, the Army embarked on an ambitious modernization

The Tank Industrial Base

The tank industrial base in the United States is an extensive network of

TABLE 1. OVERVIEW OF THE U.S. TANK INDUSTRIAL BASE

Company or Facility Location Product

Contractors to the U.S. Government

General Dynamics, Land

Textron Lycominga Stratford, Connecticut Engine

Allison Transmission Division Indianapolis, Indiana Transmission

Hughes Aircraft Corporation Culver City, California Laser range finder,

Computing Devices of Canada Nepean, Ontario Ballistics computer

FMC Corporation Alabama Track

Stanley Illinois Final drive