21 November 1986.

To: Capt. Anthony Maness Subj: Revised Comments post-Reykjavik Cf: my memo of
16 November From: J. Lederberg:

0. I very much appreciated our conversations, which helped to clarify my own thinking.
As you understood very well, I was responding - in some measure inappropriately - more
to my colleagues on the CEP than to the tasking document. Our own talks left me
fully reassured that you fully understood my intentions; but I submit this revision now
for the record.

1. The sweeping proposals for prompt total denuclearization, or for the elimination of
all OSBMs, are fraught with many flaws. To achieve those ends would require drastic
renovations of the global geopolitical atmosphere, of regimes of verification, of
conventional forces, of strategic defensive technology, that will surely take much longer
than the decade. Most constructively, we can use our imagination to outline a series of
steps dedicated to those ends, each of which can be traversed with less risk than we now
entertain. They should be more than symbolic gestures, if only to exercise the
ingredients of the renovations just mentioned.

2. The Soviets obviously will agree to nothing that does not serve their interest; and
the political dynamics of negotiation between a democratic and a totalitarian state are
likely again to give them some relative net advantage in the actual outcome of any arms
control agreement. The game is not, however, strictly zero-sum and we may have an
opportunity for substantial net gains for our own interests. We should start with
concepts that best match our security interests, shape them for feasibility on the Soviet
side, then consider the best strategies (bargains and tradeoffs) to succeed at the
negotiating table. The process is greatly complicated by inter- service competition, and
it may be awkward for the CNO to be proposing steps that seem to favor the Navy’s
preeminence in strategic forces. At least start with a clear self-understanding prior to
entering the realities of political compromise; be it with other services or with the
adversary.

3. Highly MIRV’d, fixed based ICBM’s are widely recognized to constitute the most
menacing and destabilizing parts of each side’s strategic deterrent. One of our policy
objectives should remain the elimination of these units on both sides. As the Soviet force
depends so highly on these (e.g. SS-18s) we have to consider reasonable bargaining
incentives for them to consider the costly replacement of the MIRV’da ICBMs with small
ICBMs, SLBM’s, etc. This can be mitigated by a program over a fairly extended period
of time that put the thrust of these revisions into the modernization of forces. To
sustain the verifiability of these commitments, we need early agreement to restrain the
testing of new mobile systems with MIRV capability. Our offer might also be brought
closer to parity if we brought our own land-based force to a small number, perhaps no
more than a few hundred, or even approaching zero. These would all be single-warhead
BMs. In the context of substantial net reductions, we need to do a fundamental
optimization of the mix of warheads on these, on CM’s,“bombers, on SSBNs. The least
costly option would be to retain a large part of our existing SSBN force.

4, A first step towards constructing an agreement would be a timetable agreed to by
each side for reducing the average MIRV ratio: this to be achieved by whatever mix
each side decides is in its own best interest. An early agreement to reduce this to 2.0 or
below by 1991 is probably more achievable, and may be more desirable, than an overall
force reduction that left SS-18s opposed by MXs in substantial numbers. MIRV’d
SLBMs are less dangerous than vulnerable ICBMs; but it may be difficult to persuade
the Soviets to credit this. It may even be to our advantage to trade in some numerical
disparity; e.g. to discount single-warhead fixed ICBMs in the interim counting rules (in
the same spirit as an offer to do that for bombs) but this needs deeper consideration,
and in any case is hardly to be advertised. Alternatively, an offer to take down all of
our ICBMs, large and small, might be the needed incentive for their agreement to de-
MIRV their entire ICBM force, keeping parity of total warheads. Reasonable
objections can be anticipated to an agreement that froze the US out of any future
option of small ICBMs; but this may be more tenable as an interim stage looking
towards still more comprehensive reductions as a long range objective.

5. The attached papers by May, Bing and Steinbruner show that homogeneous cuts of
up to 50% do not greatly affect the strategic balance or retaliatory capability, and
therefore the credibility of deterrence. Nor do they accomplish very much in stabilizing
the international environment, except as they indicate a trend that reverses the
alarming build-UPs of the past decade. The de-MIRVing Proposals in Section 4. are
equivalent to, or extensions of, Case C. in the May-Bing-Steibruner models.

6. The manned-bomber element of the triad stands intermediate in survivability, offers
flexibility in operations; and its potential growth may help answer the Air Force’s
concerns about sustaining its strategic role and mission.

7. Very large cuts in nuclear forces (perhaps beyond the 80% mark) begin to raise new
questions about the credibility of extended deterrence, and the adequacy of our
conventional forces. Up to that point, the contest is one of resolve more than of force.
We have little clarity how to achieve, even less how to persuasively communicate that
resolve, a persistent flaw in our models of extended deterrence: the same dilemma that
drives the aspiration to defensive approaches to security. Conventional force reduction
to one that approaches balance is a more reliable answer and may be negotiable in the
"Spirit of Reykjavik”.

8. Much of the above may soon be moot! We have not begun to face up to the
looming unverifiability of mobile BM’s and CM’s. There are proposals of marginal
credibility for verification of [production of| these systems. Time is running out for
putting such verification systems in place; and that is unlikely except in a context of far
more US/USSR cooperation than we have seen in recent history or is likely to happen.
Continued engagement with the USSR on the range of issues raised at Reykjavik is
probably the only way, if any, of evolving that context.

Our conversations also touched on needs for realistic time scales and staging of the
evolution of the demissilized world regime. At each stage, the current technology, and
forecasts for what is in the offing, need to be reassessed. For example, SDI-related
technologies may have offensive implications difficult to foresee at this time; and we do
not know just how verifiability of mobile systems will work out. On the other hand, a
strategic defense system designed to deal with a few leaks, cheats and accidents may be
achievable sooner, with more modest technology and be less likely to upset the offensive
balance than some expectations laid on the current SDI program.

Not explicitly talked about:

Enhanced cooperation in verification would be a cardinal forward step: one in
Gorbachev’s power to offer at very little tangible cost to his side would be to decrypt
their missile telemetry. Dare I suggest that this characterization of such a gesture has
any greater likelihood of success than marking the encryption as a treaty violation?
Something like that may have happened in the way the Russians are now dealing with
consultation on verification under the Biological Warfare convention. (If important, I
can amplify.)