Washington, July 12, 1971.
NSDM–117 would permit the Soviets to build up their strategic ballistic missile forces to about 2275 missiles. This figure could be higher if there is more new construction than we know about. This force would contain 288 Modern Large Ballistic Missiles. The U.S. would be limited to 1710 ballistic missiles, none of which are in the Modern Large category. This arrangement gives the Soviets an obvious and sizeable missile and payload advantage. Given the Soviet technical capability to put MRVs and MIRVs on their ballistic missiles, we face the prospect in the 1975–77 period of seeing the U.S. advantage in total warheads on target eroded. That is the last frontier of U.S. advantage in the strategic nuclear field. It could mean the end of U.S. sufficiency and parity. In addition to the inherent undesirability of this overall trend, acceding to such a USSR advantage now could be interpreted by the Soviets as compensation for forward based systems (FBS). So viewed it is a very large and unwarranted compensation.
NSDM–117 also provides for a withdrawal arrangement which would permit us to withdraw if we are unable to work out a more satisfactory follow-on agreement. I feel that this latter provision is apt to make the initial agreement practically permanent. In view of that I recommend that you reconsider the offensive proposal in NSDM–117.
One way to reduce the imbalance outlined above would be to add a provision in NSDM–117 that would require the Soviets to destroy or dismantle ICBM launchers on a one for one basis as new launchers (ICBM or SLBM) are completed. This would narrow the numerical gap between our two forces by perhaps 240 missiles. This approach does have implications for ICBM relocation and freedom to mix constraints. It also fails to remove completely the numerical gap between the two missile forces. If these implications are unsatisfactory, an earlier effective date for the freeze on ICBMs and SLBMs might be considered.