I am forwarding to you our recently completed review of nuclear targeting
policy that you requested in August 1977 as one of a number of follow-up
actions in PD–18. The study outlines
several broad policy alternatives and also makes a number of specific
recommendations with respect to nuclear weapons employment policy. It
suggests some major changes both in the thrust of current policy and in
the procedures for planning SIOP and
non-SIOP options. Some issues will
require further study and several of the broader policy issues should be
the basis for interagency discussion. In the meantime, I plan to
initiate action within DoD on those matters noted below that are within
the framework of current policy, and which I believe can be acted on
now.
The basic theme of the study is that employment policy will make its
maximum contribution to deterrence—our basic strategic objective—if our
employment policies make a Soviet victory, as seen through Soviet eyes,
as improbable as we can make it, independent of Soviet employment policy
and of any particular scenario. These plans should include targeting
options against Soviet military forces, command and control, and
military support that would maximize the threats to the objective
targets, while minimizing collateral damage. We should also have a
capability to threaten escalation. To lend credibility to a US threat to escalate, we need employment
options and supporting capabilities which the Soviets might perceive to
be advantageous to us. Such options require greater flexibility and
endurance than we now have in our nuclear posture.
As you will note, the study makes a number of specific recommendations to
these ends. In particular, it proposes specific measures to improve our
capabilities to target and attack Soviet forces, C3 and [less than 1 line not declassified] It also suggests that we
develop both plans and capabilities that will permit us to withhold
attacks on [less than 1 line not declassified]
targets as a means of coercion if a nuclear war is prolonged; and in
general, that we focus our plans and our capability on the possibility
of an extended exchange, rather than (as now) principally on an
essentially instantaneous all-out spasm exchange.
[Page 469]
It also proposes measures to strengthen
the strategic reserve forces. These measures are designed to enhance
deterrence by posing to the Soviets the prospect that they would not
only suffer severe damage in a nuclear war, but be unable to achieve
their military objectives or gain any other advantage. The proposals
also would provide a broader range of options should deterrence
fail.
Endurance is another issue of considerable importance, not only in
forces, but also in command, control, communications and intelligence.
Very few of the objectives listed above in terms of an extended
exchange, withholds, etc. can be achieved without endurance, both in
forces and in C3I. We need to consider the kind of endurance we need
and the rate at which we should proceed in acquiring a more enduring
strategic posture. Inasmuch as some of these issues relate to
acquisition policy, I have recently initiated studies designed to
identify the problems and lead to specific development and procurement
actions. In the meantime, we should take care to assure that adjustments
in targeting policy are phased to match improvements in endurance.
We have also attempted to deal with the guidelines for targeting the
Soviet nuclear threat—more effective targeting of Soviet military and
war-sustaining capacity, and targeting to threaten [less than 1 line not declassified]
In all these cases—but especially endurance and flexibility—there are
some initial steps that can be taken with the existing establishment,
but to realize the concepts fully will require acquisition decisions. In
other cases, some concepts appear promising, but require more analysis
before we can decide on concrete actions. One example of an interesting
idea where feasibility has not been determined is a “regionalizing”
strategy, to threaten continued central control of the USSR by isolating its constituents from
each other through a retaliatory nuclear strike.
I believe that the study (whose Executive Summary I hope you have time to
read) makes it clear that, while we have made substantial
[Page 470]
progress over the past year in
defining issues and proposing specific solutions, much remains to be
done to follow up on this report. Among those matters that we intend to
move on promptly within DoD are the development of the following:
We also plan to develop a program of additional studies. These will
explore such issues as how we might target Soviet general purpose forces
more effectively, [2 lines not declassified]
In addition, there are some issues that are broader in scope and,
therefore, should receive NSC
consideration. Among the most important of these is identifying and
electing a broad and coherent set of policy objectives that will give
focus to our nuclear weapons employment plans and related acquisition
policies. This report identifies four such broad policy options. It is
my view that we should adopt, at a minimum, alternative b, described on
page ix of the Executive Summary, and that we should consider moving
some distance toward alternative c, insofar as budgetary constraints
allow.
Targeting population is still another issue that is appropriate for high
level discussion. We have not in recent years targeted population per se, nor do we propose to do so now. Should
the Soviets proceed further with plans to shelter and evacuate
population, we must consider whether targeting some specific part of
population should become an explicit objective and, if so, how much of
our resources would we want to devote to that objective as compared with
other targeting objectives. Meanwhile, the NSC should reaffirm current policy.
Another important policy issue that merits attention is the development
of revised targeting plans for China. This study recommends that we
alter our targeting plans for the PRC
by handling China targeting through non-SIOP options and the Secure Reserve Force. I will proceed
to implement such a policy if you wish to direct it now; such a decision
seems appropriate.
A number of these issues should receive interagency consideration. To
that end, I propose one or more meetings of the PRC and/or SCC, prior to
an NSC meeting with you, to review this
study as well as the related ones concerning ICBM Modernization and the Secure Reserve Force.
Attachment
Paper Prepared in the Department of Defense3
Washington,
November 1,
1978
[Omitted here are the cover page and table of contents.]
NUCLEAR TARGETING POLICY REVIEW
Summary of Major Findings and Recommendations
A. Purpose
The purpose of this review was to evaluate nuclear employment
policy—that is, the policy guidelines and procedures for the
targeting
[Page 472]
of nuclear
weapons—and to identify alternatives to current policies. In
conducting this evaluation, we focused particularly on the
relationship between our stated policy (as set forth in PD–18 and NSDM–242) and the targeting plans designed to carry out
that policy. We have also reviewed, where appropriate, the
relationship between employment plans and the capabilities of forces
and supporting command, control, communications and intelligence.
The evaluation that follows uses as its framework the principal
objectives of nuclear employment policy, namely: deterrence and
essential equivalence; escalation control and war termination; and
the four general war targeting objectives described below. We also
have evaluated the Secure Reserve Force (SRF) concept, Launch Under Attack (LUA) targeting and targeting policy
for China.
B. Major Findings
1. Deterrence. Our deterrence objectives are
to deter nuclear attack on the United States; to deter attacks on
U.S. forces abroad and on our allies; and to impede coercion by
unfriendly nuclear powers of the US, its allies and other friendly
nations. Nuclear weapons play a major role in meeting these
deterrence objectives, but they are not expected to do this task
alone.
While we are not sure what deters the Soviets, there is fairly broad
consensus in the US intelligence
community and among a number of Soviet experts that the Soviets
seriously plan to face the problems of fighting and surviving a
nuclear war should it occur, and of winning, in the sense of having
military forces capable of dominating the post-war world. Their
emphasis on planning for nuclear war and on damage limiting
measures, including civil defense and civil emergency preparedness
testifies to this overall thrust in Soviet policy. This does not
mean that the Soviet leadership is unaware of (or indifferent to)
the destructive consequences of a nuclear conflict. Indeed, there
are many statements by Soviet leaders which attest to their desire
to avoid nuclear war and to their recognition of its potential
destruction. However, the Soviets appear to have prepared themselves
militarily and psychologically for the possibility that a nuclear
war could occur and within the limits of their resources, they have
prepared plans and developed capabilities which would permit them to
do as well as possible in surviving a nuclear conflict and in
defeating the military forces of their adversaries. It is clear that
they are continuing substantial efforts to improve their own
strategic posture. The effect is to pose new obstacles to
achievement of our strategic objectives. We do not argue that the
US concentration on deterrence
is wrong, or that the Soviet idea that nuclear wars are winnable is
right, but rather that carrying out a policy of deterrence cannot
ignore these Soviet attitudes.
[Page 473]
Our deterrent appears adequate in normal circumstances to prevent the
Soviets either from attacking us or our allies deliberately or from
pursuing a recklessly aggressive policy carrying with it a high risk
of war. But it is in a severe crisis that our ability to impede
coercion and extend deterrence to other nations would be most
severely tested. Should such a crisis occur, we would want to avoid
war (or terminate a war at the lowest possible level of violence)
while simultaneously preventing the Soviets from coercing us or our
Allies. In such a case, deterrence requires that the Soviets must
never be confident that escalation would be to their advantage; also
they must never be certain that the U.S. is unwilling or unable to
respond effectively to any attack.
Deterrence will be influenced primarily by Soviet perceptions of our
capabilities and will, rather than our plans. However, to the extent
that our plans are known to the Soviets, these plans say something
about our capabilities and will. Employment policy also has an
impact both on the Soviet perception of the risks and advantages if
they escalate, and on our confidence which in turn affects the
Soviet perception of the likelihood that we may escalate.
Since the Soviets appear to have a concept of military victory, even
in nuclear war, we should seek employment policies that would make a
Soviet victory as seen through Soviet eyes, as improbable as we can
make it in any contingency. Thus, we should develop plans and
capabilities that minimize Soviet hopes of military success. These
should include targeting options against Soviet military forces,
command and control, and military support that would maximize the
threats to the objective targets while minimizing collateral damage.
We should also have a capability to threaten escalation ourselves.
This threat to escalate if and as necessary is at the heart of
NATO’s flexible response
strategy. It is likely to be especially effective if it threatens
Soviet ability to maintain effective military forces in the
field.
It is sufficient for purposes of deterrence if the Soviets perceive
that there be a reasonable likelihood that we could (and would)
escalate or respond successfully; it is not necessary that we have
highest confidence that escalation
control will work, or, still less that we can win the war. However,
to lend credibility to a U.S. threat to escalate, we need employment
options and supporting capabilities which the Soviets might perceive
to be advantageous to us. Such options require greater flexibility
and endurance than we now have in our nuclear posture.
2. Escalation Control. There are and will
inevitably always be great uncertainties about our ability to
control escalation and terminate conflict on terms acceptable to us
and our allies. Nevertheless, we conclude that it remains in the
U.S. interest to have plans and capabilities that could limit damage
by controlling escalation and terminating a conflict before it can
extend to all-out nuclear war. Thus, we reaffirm the
[Page 474]
desirability of a policy of escalation
control based on a range of SIOP
and non-SIOP options. We find,
however, that there are serious deficiencies in current plans and
capabilities to carry out a strategy of escalation control. There
has been inadequate political input into the planning of nuclear
options, particularly non-SIOP
options. There are deficiencies in the integration of limited
nuclear options with non-nuclear plans, and an absence of political,
economic and psychological plans to complement non-SIOP options. Further, the plans for
limited use of nuclear weapons have not been sufficiently exercised
with the participation of high level political and military leaders.
The vulnerability of forces and Command, Control, Communications and
Intelligence (C3I) also limits the effectiveness of an
escalation control strategy. As a result, the US Government may not be adequately
prepared to deal with a crisis which could involve nuclear weapons,
should it occur. Dealing with a nuclear crisis in the multilateral
framework of NATO would be even
more difficult.
3. General War Plans. Our general war plans
are designed to meet the following principal objectives, the last
three of which are to be accomplished “to the extent practicable:”
(1) impede recovery of the Soviet Union both in the short term and
the long term; (2) destroy Soviet national political and military
leadership and command and control; (3) destroy Soviet nuclear
forces, and (4) destroy Soviet non-nuclear forces.
Although targeting to impede recovery receives
highest priority [less than 1 line not
declassified] it is not clear that
threatening to impede recovery by destroying large amounts of
Soviet population and industry is the most effective deterrent,
particularly in situations less than general war. Nor is it
clear that our current targeting would, in effect, subsequently
impede recovery, in the long-term (as distinct from reducing Soviet
GNP sharply, which it clearly
would do). Furthermore, we have no confidence that our present
targeting plans would prolong Soviet recovery more than our own if
massive attacks were launched by both sides. While planning
contemplates the possibility of withholding attacks on recovery
targets if substantial US
urban/industrial assets survive an initial attack (and this is
appropriate in our view), the endurance and survivability of our
forces and their supporting command, control, communications and
intelligence are not sufficient to support such a strategy.
In all large scale attacks on Soviet recovery targets (and indeed on
military forces) there would be substantial Soviet population losses
(at least tens of millions). But, if Soviet civil defense plans are
successfully implemented, these levels could be significantly
reduced. If the Soviets or the US
could effectively shelter and thereby protect a significant portion
of their labor force, this should have an important bearing on
[Page 475]
recovery, for the
surviving skilled labor force will be an important element in
achieving recovery.
[13 lines not declassified] Under present
plans the attack on political leadership would involve substantial
collateral damage to the general population assuming that the
population is unsheltered and unevacuated.
[1 paragraph (10 lines) not declassified]
The U.S. targets the Soviet nuclear threat to
achieve two objectives: the first is limiting damage both to the
U.S. and our allies; the second is to prevent the emergence of a
post-war nuclear balance that would facilitate coercion by the
Soviet Union. It also is apparent that the criterion for
destruction of Soviet nuclear capabilities, i.e., “to the extent
practicable with available allocated nuclear forces” is extremely
general—hardly a precise guideline for target planners. Moreover,
there are substantial uncertainties associated with this objective
(as with others). [5 lines not declassified]
One fact is clear, however. The proliferation and hardening of
Soviet missile systems have substantially eroded our counterforce
capabilities over the past decade. We have found no plausible
changes to targeting policy or force structure in the course of this
study that give any promise of restoring the relative capabilities
we enjoyed in the early 1960’s. This does not mean, of course, that
we can or should do nothing to improve the present or prospective
balance. Cruise missiles will put Soviet land based missile systems
and other hard targets at risk again, but this will not give us a
prompt capability; MX and TRIDENT II will, however.
With respect to the damage limiting objective,
today there are two distinct views as indeed there have
been for some time. One view holds that since we cannot expect to
limit to low levels the damage resulting from a large scale nuclear
attack, that it is no longer a meaningful objective and should be
abandoned or at least given a low priority in employment policy. A
central concern is that continuance of damage limiting as a major
objective of U.S. policy could lead to increased arms competition
without any resulting improvements in U.S. security and could divert
forces from more promising objectives. The opposing view is that we
must continue to do the best that we can to protect the U.S. from
the consequences of a nuclear war if deterrence fails. Given the
uncertainties of nuclear war, and the wide range of possible
scenarios, there might well be situations where the capability to
reduce damage by perhaps tens of millions of American lives would be
far from futile. This view also stresses the potential effects on
deterrence and crisis management in situations short of nuclear war
if U.S. society were to become far more vulnerable than the Soviet
Union.
[Page 476]
With respect to the objective of preventing an
unfavorable postwar nuclear force balance, the debate turns on
what constitutes a balance and on the best means for achieving
it. Recent changes in the strategic balance pose us with a
dilemma—how much of our force should we use in an effort to erode
the Soviet nuclear threat and how much do we hold in reserve to
secure a post-war balance? A substantial portion of the forces
available for SIOP [less than 1 line not declassified] is used
for the counter-nuclear mission even though relatively low damage
expectancies are anticipated. Allocating additional weapons with
current capabilities would not be productive. Indeed a considerable
number of the weapons now used are SLBMs which have low DEs [less than 1
line not declassified] against hard Soviet missile silos.
Furthermore, as ICBMs, which have
a better hard target capability, become more vulnerable to attack,
withholding them for use in other missions may only result in their
loss. Some argue that we should allocate only weapons with the best
hard target capability to the nuclear threat and not allocate SLBM weapons with low DEs to hard
targets. Others argue that the present scheme of cross-targeting is
a hedge against failure of one leg of the TRIAD, and that given
uncertainties as to what actually would kill a silo (or interfere
with reload and force reconstitution), we should continue to
allocate [less than 1 line not declassified]
to each silo, (even if some have low DEs against the silo itself) at
least as long as we have sufficient weapons.
Because a substantial portion of the Soviet nuclear
threat is hard, a major issue is how much and what kind of
capability is required for this targeting objective. This
is an area in which programmed force changes—and acquisition
decisions—and employment policy are closely linked. For example, the
large ALCM force that will be
available by the mid-80’s will have the accuracy for a considerable
counter-silo potential if ALCM
penetrability is high. The US also
faces basic decisions about the characteristics of future ICBM and SLBM forces in which the requirement for Quick Reaction
Hard Target Capability (QRHTC) is
a driving factor.
From the standpoint of targeting it seems clear that we ought to
retain a substantial hard target capability. Such a capability is
required not only to be able to attack Soviet ICBM silos effectively, but also for
the growing number of hardened C3 facilities and some other
hardened installations. Improved HTC would enable us to reallocate weapons with low DEs
against hard targets to other missions. Whether a substantial
portion of our HTC needs to have a
rapid response capability cannot be determined on targeting
considerations alone. The targeting requirements for rapid responses
are very scenario dependent. For example, if Soviet forces are alert
when the US launches a counterforce
attack the probability of their preemption or launch under attack is
high, and the difference in response time between a few hours and
[Page 477]
a few minutes may be
inconsequential. On the other hand, given the many uncertainties
noted above, a quick hard target capability might well improve the
outcome of a nuclear exchange from our standpoint or complicate
Soviet calculations of the outcome and thereby help to strengthen
deterrence.
We have also found that with current plans, attacks
against Soviet non-nuclear military forces are likely to be
ineffective in many scenarios. The current set of targets
attacked in the other military targets category, includes only fixed
installations. Our knowledge of Soviet war plans suggests that with
plausible amounts of warning, both forces and stocks would be
dispersed rapidly away from these fixed installations. Thus, much of
this attack, unless the US achieves
total surprise, is likely to go on empty or partially empty kasernes
and other bases, and the Soviets would be left with substantial
military power to coerce other nations, to seize valuable industrial
resources in Western Europe and the Middle East and to assist them
in post-war bargaining with the United States. On the other hand,
there are clearly a number of fixed military installations that will
remain valuable in supporting the Soviet war effort, regardless of
warning. For example, secondary airfields, transportation centers,
etc. Current planning does not give these targets high priority in
relation to facilities that are likely to be evacuated with warning.
We believe future planning should take this into account. For the
longer run, priority should be given to capabilities to attack
dispersed military forces. Trans-attack reconnaissance and
responsive targeting are needed to do this.
With regard to the strategic reserve force we find
that the force and its supporting command, control,
communications and intelligence (C3I) may need to survive for weeks or
even months after an initial nuclear attack. These elements
lack sufficient endurance today to meet such an objective. If the
SRF is to meet its stated
objectives, the principal criteria for composition and sizing should
be its endurance and its relative capabilities in relationship to
Soviet plans for secure reserve forces. Indeed, many of our
requirements for more flexible and discrete targeting at lower
levels also imply larger reserve forces. The force mix in the Secure
Reserve Force also needs reexamination to assure that it has maximum
endurance. C3I supporting the reserved forces also needs
greater endurance. Furthermore, the current provisional target sets
for the Secure Reserve Force (primarily low-priority economic
targets) do not contribute significantly to the objective of
post-war coercion.
In relation to NATO our deterrence and escalation control
objectives require an effective NATO employment policy. Allied concerns about
our ability to deter aggression against NATO have grown as Soviet capabilities have grown. More
effective plans for targeting the Warsaw Pact threat to NATO with strategic weapons and
closer coordination of US
[Page 478]
and NATO planning could help to alleviate
these concerns. However, to make such plans effective requires more
responsive and survivable C3 and intelligence, and even
closer integration of employment planning and crisis management
between the U.S. and its allies at both the military and civilian
levels. For the longer run, the availability to SACEUR of an option to target a
full range of threats to Allied Command Europe (ACE) without
invoking the SIOP would also
strengthen deterrence and the confidence of our allies.
There is no plan at present for launch under attack
of only the ICBM force.
If the ICBMs were launched against
their current SIOP targets there
would be substantial collateral damage to Soviet population and this
would likely invite retaliation against US urban/industrial assets. In any event, the set of
targets for our ICBM force is not
in itself planned to meet any specific objective. We ought to have
an option to launch only our land-based ballistic missiles against a
target set which would serve some defined objective while minimizing
collateral damage, thereby reducing the prospect of retaliation
against a broad US urban/industrial
target set. The launch under attack doctrine should not be seen as a
solution to the problem of ICBM
vulnerability. In many situations, LUA, would, in effect, rule out a real choice by the
President. The President should not be forced to make a “use or
lose” choice if there are other options available. Reliance on
LUA increases the risk of an
accidental war and thus would increase instability in a Soviet-US
crisis. However, we do believe that targeting plans should include a
LUA option for ICBM forces only that is less
escalatory than current plans.
The current SIOP
targeting policy for China is out of date. It was based on
a period when China was seen as a threat comparable to the Soviet
Union. We believe that our China targeting policy should be
reformulated to bring it more into line with current political and
military realities. China poses no strategic nuclear threat to the
US today and will only have a
minimal capability within the next several years at least. At
present, US-Chinese relations are improving, and the PRC is clearly more menacing to our
adversaries than to our allies. Political relations could change,
but we would likely have a good deal of warning. In any case, while
it is not clear what will deter China, it seems unlikely that a
primarily agrarian society with a small industrial base will be
deterred from regional aggression against its neighbors by the
threat of massive attacks on industry. We not only do not understand
the recovery process as it relates to China, but we are quite
arbitrary in assigning value to those targets we select. The current
requirement to program at least one weapon on an industrial facility
in the top 125 urban areas in the PRC, drives, to a large extent, the high weapons
requirements for China targeting even though over 50 percent of
China’s industry is located in 25 cities.
[Page 479]
C. Major Policy
Alternatives
We have developed several alternative employment policies that we
believe, on the basis of our study, to be representative of the
choices facing national leaders with respect to future employment
policy. The identification and assessment of major policy
alternatives is a somewhat artificial exercise. If precedent is any
guide, policy is more likely to be determined incrementally by a
series of discrete decisions about what to procure and when, how to
phrase a given policy statement, what to include in an arms control
proposal, etc. Thus, there are, in actuality, a number of choices
that could be made. Nevertheless, it can be useful to identify and
assess broad policy as a framework for making more specific
decisions. And such decisions should, if possible, be made with some
set of overall objectives in mind.
There are several factors that are likely to dominate the choice of
alternatives. Most important are assessments of Soviet views and
objectives with respect to the role of nuclear weapons; and in light
of these Soviet views and objectives, judgments as to what actions
we should seek to deter with nuclear forces, and how best to do so.
In this connection we also need to consider: (a) what flexibility in
our nuclear posture (i.e., how broad a range of options) is desired
and what is feasible and how much we should spend on it; (b) how
much endurance do our forces require and how much is possible; (c)
how much damage limiting capability is considered necessary and how
much is possible; and (d) the costs of achieving these capabilities.
One alternative, of course, is to continue current policy as
described above. We offer below four broad alternatives to current
policy.
All of the following options will meet the current objectives of
nuclear policy to some degree. However, they differ in their
relative emphasis on flexibility, endurance and counterforce; and as
a result could have substantially different cost implications.
However, each contains at a minimum, an assured destruction
capability.
a. One alternative is to strengthen current policy,
particularly by improving the flexibility of plans and the
endurance of forces and their related command, control and
intelligence. Under such a policy, an assured capability to
destroy industrial targets of value to the Soviet Union would remain
the backbone of deterrence, and would receive emphasis in
declaratory policy. However, the goal of “impeding recovery” would
be redefined to focus [2 lines not
declassified] The forces and related C3I to
accomplish this mission should be given additional endurance over
time so that attacks on industrial targets can be withheld so long
as substantial U.S. industrial value remains undamaged. Added
emphasis would be given both in planning and declaratory policy to a
more effective means of targeting Soviet conventional forces and
command and control as a supplement to assured destruction of
industry. Counterforce objectives
[Page 480]
would deemphasize damage limiting and focus
targeting on preventing the emergence of marked asymmetries in
US and Soviet capabilities that
could be exploited by the Soviet Union to coerce us or our allies
following a nuclear attack; or which, if perceived as an advantage
by the Soviets, could affect crisis bargaining short of nuclear war.
This policy would also retain non-SIOP options, but strengthen the procedures to
integrate non-SIOP nuclear options
with other military and political measures. This policy would
involve alteration of current targeting plans and declaratory policy
in order to take into account what we know of Soviet views of
nuclear strategy, in particular their sensitivities to losing
control over their society, and the deterrent effect that we might
achieve by planning to attempt to deny the Soviet Union a war
winning capability. Some believe such changes to current policy
represent the minimum necessary to strengthen deterrence in light of
what we know of Soviet objectives and their growing military power.
Others believe that changes to current policy are not necessary to
strengthen deterrence and would be provocative and costly.
b. A further departure from current policy would be
to focus both employment and declaratory policy more heavily on
denying the Soviets any confidence of achieving a favorable war
outcome. A high priority effort would have to be put on
developing greater endurance and on improved targeting of [less than 1 line not declassified] New
capabilities would be required to support such a targeting policy,
particularly more survivable C3I. Countervalue targeting
would (as in a.) focus on Soviet [1 line not
declassified] However, these attacks would be designed to
be withheld for [less than 1 line not
declassified] as part of an expanded reserve force concept.
Targeting population or targeting industry to impede long-term
recovery would not be a specific SIOP objective, but an assured destruction capability
(to be withheld so long as the Soviets spared U.S. cities and
industries) should be maintained. Targeting of both Soviet nuclear
and conventional forces would be designed principally to assure that
they could not expect to achieve a favorable nuclear or military
balance following a counterforce exchange. However, damage limiting
to the extent feasible would also be retained as an objective of
counterforce targeting. Non-SIOP
option planning would be improved as in a. above. Declaratory policy
would stress that the overall objective of our nuclear policy is to
deny to the Soviet Union a favorable outcome from a nuclear war.
Some argue that this policy could help to convince the Soviets that
the US was seeking serious
war-winning objectives (as some maintain the Soviets do) and thus,
would be a far more effective strategy in extending deterrence and
preventing coercion, and that it would give the U.S. more reasonable
war objectives if deterrence fails. Others argue that by reducing
emphasis on the prospect of massive retaliation and by implying that
the U.S. was
[Page 481]
more prepared
to fight a nuclear war than we have been in the past, we would
weaken deterrence and decrease stability. This policy could prove
more costly than Option a., but whether it would be substantially
more costly is not certain.
c. Still a further departure from current policy
would add a higher confidence capability to limit damage.
This would require not only greater capability against Soviet
nuclear forces than in Option b. above, but also substantial
improvements in defenses. Under this policy, we might return to the
targeting objectives of the earlier SIOPs. For example, we might attempt to achieve
something like [1 line not declassified]
threatening the U.S. and our allies under all circumstances of war
initiation. Obviously, the forces required for such a strategy would
be substantial and would have to be acquired over a period of years.
Thus, this could not be a short-term objective of U.S. policy. Some
would argue that a damage limiting capability at least comparable to
that of the Soviet Union is the sine qua non
of essential equivalence and a necessary requirement to make a
strategy of escalation control credible. Without the ability to
deter escalation at the higher levels the U.S. could not count on
controlling escalation at lower levels. Others argue that the
achievement of such a damage limiting capability would be highly
destabilizing and would hardly be feasible given the Soviet
capability to respond to whatever measures the United States might
take. It seems clear that this option would be more costly than
current policy or either of the two prior options.
d. Finally, the U.S. might choose to move to the
other direction from current policy and rely more heavily on
assured destruction defined in either terms of industry,
population or cities. This would avoid the need, perhaps
quite costly, to improve current deficiencies in flexibility and
endurance. Moving in this direction would imply a judgment that the
post-war nuclear force balance is not a meaningful measure of
“victory” and that the prospect of massive destruction is a credible
deterrent for large scale attacks including those aimed at nuclear
forces. A continued capability to execute a wide range of limited
attacks would be possible with the forces provided by this approach,
but, in general, the approach assumes that any nuclear exchange is
likely to escalate very rapidly to all-out countervalue exchanges.
Indeed, it depends on that prospect for its deterrent effect. The
argument against this approach is that it would narrow the scope of
deterrence. In particular, such a policy would have an adverse
impact on extended deterrence and thus on alliance relationships,
and might suggest opportunities in the future for the Soviets to
utilize their nuclear forces for coercion of the US and our allies. It would provide the
US with a very narrow range of
options should deterrence fail.
[Page 482]
D. Major
Recommendations
1. Greater flexibility should be built into the
SIOP through the
development of more discrete building blocks which could
break down the present target base into smaller increments and
thereby give the President a broader range of options if he should
ever have to consider SIOP type
attacks. Each building block should have distinct targeting criteria
which take into account not only the timing and damage requirements
for attacking the objective target but also collateral damage to
other target sets. Given the planning complexities, the development
of building blocks requires an evolutionary approach with close
interaction between policy levels and planners. Care must be taken
to insure that a balance is struck between the quantity of useful
options desired and the need to maintain a relatively simple and
responsive execution process. (See Issue #1)4
2. The requirement for endurance should be
considered a high priority requirement in the future planning of
US forces, command, control
communications and intelligence assets. Endurance—the
ability of strategic nuclear forces not only to survive the initial
attacks but to remain an effective military force for a prolonged
period afterwards—is a key element in any strategy that pays
attention to post-exchange balances and/or the possibility of a
drawn-out series of exchanges. Specific recommendations for
endurance measures involve acquisition policy, and thus are beyond
the scope of this study. However, we can say that to carry out
current employment policy effectively, much less the more demanding
alternatives outlined above, the endurance of command, control,
communications and intelligence assets need to be improved
substantially so as to make it possible to support the concept of a
Secure Reserve Force and withhold attacks (e.g., on Soviet
non-military industry) so long as substantial US urban/industrial assets remain
undamaged. It is important that modifications in employment policy
and plans that rely on greater endurance proceed in phase with the
improvements in our force posture and supporting C3I that
are necessary for endurance. (See Issue #2)5
3. We should target Soviet nuclear forces and
develop our own forces so as to maintain roughly equal
counterforce capabilities. In particular, counter-nuclear
targeting should be designed primarily so as to assure that the
Soviets are unable to shift the balance of nuclear power
drastically by attacking our forces, and so that it is clearly
perceived they cannot. This objective cannot be achieved
solely by attacking Soviet forces and thus is not solely a function
of targeting policy. It also requires that we be able to hold in
reserve forces comparable to reserved Soviet forces, so as
[Page 483]
to prevent post-war
coercion and thus protect these forces and their related C3I. While
we lack the ability to limit damage to the US society meaningfully in a large scale Soviet attack,
we are reluctant wholly to eliminate this as an objective of US policy, particularly because to do
so explicitly would appear to confirm a major asymmetry between
US and Soviet policy and would
ignore important uncertainties about the effectiveness of a damage
limiting strategy. However, if we focus on avoiding asymmetries in
nuclear power in developing targeting plans, we are also likely to
do as well as we can expect to do in limiting damage.
We recommend the following specific guidance for
targeting the Soviet nuclear threat to the US and our allies. First, we
should, for reasons of alliance solidarity, continue to give equal
priority to targeting threats to the US and to our NATO
allies. Second, when forces are fully generated and there are
sufficient weapons available to meet other targeting objectives
adequately, including the maintenance of a Secure Reserve Force, we
should continue to plan to place at least one weapon on each target
that constitutes a nuclear threat to the US or our allies, using the most effective weapon for
each type of target and taking into account the desirability of
promptly using forces with less endurance. For the longer run we
should have sufficient weapons with hard target capabilities to meet
this objective without utilizing weapons with low PK. Third, in the
retaliatory case, priority should be given (among nuclear threat
targets) to [6 lines not declassified] It
will be important in the longer run to improve the US capability to acquire information
rapidly on the status of Soviet strategic forces following an
attack. (See Issue #3)6
4. New priorities should be established for
targeting Soviet non-nuclear forces taking into account the
probability that Soviet forces will be dispersed upon warning. Any
victory-denial approach should pay close attention to the ability of
nuclear weapons to affect the post-exchange balance of military
forces, broadly defined, not just nuclear forces. This will require
a special effort to identify the [1 line not
declassified] does not decrease greatly with warning, to
include [1 line not declassified] The target
value system would be adjusted to assure the destruction of these
targets and to give [1 line not declassified]
For the longer run, we should initiate a high priority special study
on the feasibility of targeting [1 line not
declassified] requirements for accomplishing this.
Particular attention should be given to how strategic forces might
be utilized more effectively in support of NATO. A target package should be developed to [1 line not declassified] (See Issue #4)7
[Page 484]
5. Targeting of the Soviet [5 lines not declassified] A high priority effort should be
undertaken to identify and target [3 lines not
declassified] We should continue to have an option to
withhold attacks [4 lines not declassified]
should receive further study on a high priority basis along with
[less than 1 line not declassified] (See
Issue #5)8
6. Targeting of Soviet [2
lines not declassified] This attack should be designed so
as to minimize collateral damage to population [less than 1 line not declassified] consistent with
achievement of the attack objective. It should be possible to carry
out this attack [1 line not declassified]
Second, [4 lines not
declassified] by the US
during and after the war.
We recommend that the US continue current policy with respect to the
targeting of population, in which population, as such, is not an
objective target. At the same time, we recommend continuing
to plan. [2 lines not declassified] Unless
Soviet civil defense becomes far more effective than presently
estimated, there will be substantial population at risk in any such
large scale attack, as is the case now. We find no reason to believe
that targeting population per se, would be a
more effective deterrent or a more useful objective in general war
than targeting the specific economic objectives suggested above
along with the control apparatus and military power which the
Soviets appear to consider of high value. Furthermore, targeting
population would require substantial additional allocation of
weapons if we assume that the Soviet civil defense is implemented
and effective, and therefore would divert weapons from other
objectives. However, estimates of population fatalities will
continue to be an important criterion for any decision maker
contemplating the use of nuclear weapons. Our data and methodology
for making such estimates should continue to be improved. We should
also keep under continuous examination the feasibility and the
implications for other targeting objectives of adjusting our
targeting so as to be able to attack some defined portion of Soviet
population even if it is evacuated and/or sheltered. Whether we
should have a specific target set for use in such a case remains an
unresolved issue.
7. We should continue to plan a Secure Reserve
Force (SRF) as part of the
strategic reserved forces. Our long run objectives should
be to assure that reserve forces in a prolonged nuclear war at any
stage of that conflict would be superior or at least comparable in
capability to the forces of the Soviet Union. A principal objective
of the Strategic Reserve Force would be to deny the Soviet Union the
possibility of changing the correlation of forces. We need to
consider the role of theater-based forces in a secure reserve
concept and the possibilities for augmenting forces during a crisis
or after a limited attack. We need to develop capabilities that
would permit flexible retargeting of the reserve force for we see
[Page 485]
no way in which this
force can be realistically pretargeted prior to a nuclear
engagement.
For the short run, we recommend that the
composition of the Secure Reserve Force be based heavily on the
probability of survival and endurance in its components. Given this
concept, the principal objective of the Secure Reserve Force should
be to achieve enduring survivability. What it is targeted against is
less important than its ability to survive and endure. However,
during the period when we lack an enduring intelligence and
retargeting capability, the Secure Reserve Force should continue to
have tentative targets likely to have high continuing value even
after an initial attack, e.g., bomber bases. This would permit,
under worst circumstances, follow-up strikes to be executed “in the
blind” against targets likely to have continuing value to the Soviet
Union. (See Issue #7)9
8. In addition to developing more discrete SIOP options, the process for the
planning and use of non-SIOP
options should be improved. The only way to develop
realistic political/military contingency plans is through a
continuing interactive process between the planner and the
policy/decision maker. It is in the nature of limited nuclear
options that there will be a high political input into any
consideration of the use of such options. And, if they are to
achieve their stated objectives, the other associated military and
political measures must be closely integrated with the use of
limited and regional nuclear options. We, therefore, recommend that
the current planning process be modified to include regular
interaction between the JCS and
OSD, including selective
representatives from the State Department, the NSC staff and the Intelligence
Community. Planning should include periodic exercises to test both
the feasibility of implementing the plan and to expose
policy/decision makers to the plans and give them an opportunity to
evaluate them under as realistic conditions as possible. While DoD
cannot organize the crisis management machinery of the US Government unilaterally, the SecDef should recommend to the
President’s National Security Advisor the development of a national
crisis management mechanism based on the planning procedures
described above. (See Issue #8)10
9. Closer coordination of nuclear planning between
US and NATO planning staffs should be
undertaken. In particular, USCINCEUR and CINCLANT should be
encouraged to develop additional US
employment options in support of SEPs. Closer integration of nuclear
planning between US and NATO planning staffs is needed,
particularly in the development of target plans which integrate
theater and strategic
[Page 486]
nuclear forces in striking the Warsaw Pact nuclear and conventional
military threat to NATO. Should
further analysis suggest that more effective ways can be found to
target the Soviet/Warsaw Pact threat to NATO, these should be the basis for discussion with our
allies under the aegis of the NPG. The opportunity should be seized
to involve the allies in a more meaningful way in the development of
alliance nuclear employment policy. Improvements in the vulnerable
and outdated NATO
C3
system clearly are needed, along with better means for rapid
processing of intelligence. We need to take steps that assure that
both NATO Selective Employment
Plans (SEPs) and SACEUR’s
Nuclear Operations Plan (NOP) are consistent with our own; that if
current LNOs or SAOs are employed there will be
corresponding NATO plans that are
complementary rather than conflicting. (See Issue #9)11
10. The JCS should
develop a launch under attack package for ICBMs only that will be
directed at a range of military and defense production targets but
will result in minimum collateral damage consistent with achievement
of its targeting objective. This launch under attack package should
be ready for use beginning in the 1981–82 period and should include
a broad set of nuclear and non-nuclear targets and command and
control. It should also include such targets as the Soviet ASAT launch facilities and Soviet
ASW bases which might support
attacks which could reduce US
endurance. The attack should be designed so as to minimize
collateral damage to population consistent with achievement of the
attack objective. We do not see LUA
as a solution to the problem of ICBM vulnerability, but believe such an option should,
nevertheless, be available to the NCA. (See Issue #10)12
11. Employment policy for China should not require
the extensive planning process which is devoted to the Soviet
Union. We should, of course, recognize that China does pose
a threat to some US interests in the
Far East and that the PRC might
attempt to coerce US friends or gain
assets of interest to us, particularly in the aftermath of a
US-Soviet exchange. We would assume that if China’s posture
substantially changed, we would be sensitive to this and could
accommodate modifications in our targeting policy accordingly.
Implicit in this recommendation is the belief that U.S. and Allied
conventional and U.S. theater nuclear forces (using non-SIOP options) are sufficient to deter
the likely range of Chinese threats in peacetime and that the SRF, available for protection and
coercion worldwide could be used to deter China in a trans- and
post-attack environment. (See Issue #11)13
[Page 487]
12. The data base for targeting needs to be revised
and expanded. It is evident from past experience that the
design and maintenance of a responsive target intelligence data base
is very complex and any change in policy portends significant
modifications in data. For these reasons, we recommend JCS evaluation of the impact that the
targeting policies contained in this report will have on the ability
to produce and maintain an adequate, comprehensive, responsive
target intelligence data base. The JCS should provide a plan, with appropriate milestones
and resource requirements, to provide a flexible data base.
13. The development of nuclear employment policy is
an ongoing process that requires continuing interaction between
policy makers and planners. Presently, there is no
mechanism or arrangement that could assure that our employment
policy is developed on this basis. The Secretary of Defense should
create within OSD a mechanism to
conduct the necessary follow-on work and assure its continuity.
[Omitted here is the body of the report.]
