Why do we need an MX? The stock answer is because Minuteman's survival is at stake. But if one thinks about it, building the MX really doesn't help Minuteman. It would simply provide (assuming a limit on the number of Soviet and Chinese and French and British reentry vehicles) that if Minuteman were to be targeted and attacked, MX might survive for a retaliating strike -- but not Minuteman . Minuteman is part of our triad of strategic systems whose function is to maintain a credible deterrent. If we have to build MX so that we can fire it after Minuteman has been destroyed, we haven't helped Minuteman -- and we certainly haven't maintained a deterrent, so we really haven't gotten much for our $100 billion.

One of the first decisions a president should have to make relates to whether he'll ever order a retaliating ICBM launch against a foreign power. If the answer is no, then we really don't need an MX or anything else during his tenure in office. But we certainly should not let that fact be known. If the answer is yes, then the question is when and under what conditions would he launch such an attacak. One could ask: "Would he retaliate after receiving 100 foreign detonations on the United States?" He would probably say yes. The next question should then be: "How soon after and at what?" He would probably say: "Immediately and at the original targets." By this questioning, we have narrowed down the decision time to something like a half-hour -- the time between the launching of the foreign missiles and their arrival.

Then this question might be appropriate: "Mr. President, would the enemy warheads have to hit the ground before you would launch?" How a president would answer that question would depend on his philosophy with regard to nuclear weapons but, more important, on the accuracy with which one could guarantee that the information being supplied to him was valid.

Technology is available today to determine whether an appreciable fraction of Minuteman is at risk from any foreign ICBM launch. The key works here are "appreciable fraction." Important to this determination, however, would be the ability to guarantee that the information received is valid and cannot be interrupted. Instead of having a minimal number of satellites and a few ground stations -- with sometimes cantankerous computers -- monitoring foreign ICBM launchings, one would need redundant systems numbering in the tens. Many of the satellites would have to be silent, only talking to each other and being appropriately monitored. Some be decoys, but I suspect that the launching cost would dictate that all space hardware be operable. The shuttle could be the answer to cutting costs in putting this hardware in space.

Such redundant monitoring systems could provide absolute information about the risk to Minuteman. If Minuteman were at risk, it could be launched so that it would no longer be at risk. With proper information, Minuteman could be launched at other-than-empty foreign silos. Moreover, the whole Minuteman force would be available. A properly designed system could serve as a much more credible deterrent than the MX, which would always be vulnerable to an increased number of foreign re-entry vehicles, which could be built in a shorter period than could the MX. Such a system would be cheaper, could be in place sooner and would help put a cap on the arms race because not one new offensive system would be required. It would also greatly improve the present alerting situation, where the failure of an elelctronic component causes local and international concern.

When the concept of being able to launch after attack is suggested, many individuals believe one is proposing an automatic system. Not so. The system, as with all of our nuclear missile systems, could be launched only upon the president's orders. It should be no more automatic than any existing system. It would, however, have the advantage of providing the president with credible, advanced information with which he could make a decision. He would not have to wait to have a sizable fraction of Minuteman or any other part of the United States destroyed before he could be certain that we were under a real attack. Such a system could guarantee that Minuteman would never be at risk if the president did not wish it to be at risk. This, to me, would maintain a very credible deterrent at a much lower cost than any MX would require.

If 1,000 non-vulnerable Minuteman missiles are a sufficient deterrent, additional MX missiles based in Utah would not be required. This is not to say that the Minuteman missile should not be replaced by the MX missile. As Minuteman ages, it must be replaced, and the MX is a logical candidate. And if, alas, a completely new MX deployment mode is required, the information that a redundant satellite alerting system would provide could be available to reduce the vulnerability of even that system. Clearly, Minuteman, or preferably the MX missile in the Minuteman silo, would need to be appropriately hardened so that it could indeed be launched when threatened.

Serious consideration should again be given to an active hard-site defense of Minuteman. Compared with the weaknesses of the present proposed MX deployment, this option should be given much greater consideration than it has in the past. My proposal lends itself better to the role of a meaningful deterrent as well as to future arms control options, since the number of strategic systems initially would not be increased, and the present full effectiveness of our strategic deterrent could be preserved.