Two Witnesses for the Defense
APPROVED FOR RELEASE 1994
CIA HISTORICAL REVIEW PROGRAM
18 SEPT 95
An analyst view on the problem of the clandestine collector.
TWO WITNESSES FOR THE DEFENSE
Harlow T. Munson and W. P. Southard
We have read Mr. Rumpelmayer's statement of his reservations1 and we feel both qualified and obliged to offer some testimony. We were the principal officers of two CIA groups which spent a year -- working separately on complementary studies -- making reconstructions of the Soviet venture in Cuba; neither of us had been previously engaged with Cuba. Our two studies,2 which considered the same range of questions but different bodies of evidence, arrived at similar conclusions, which are far from Mr. Rumpelmayer's.
Mr. Rumpelmayer's basic contention -- the burden of his articles -- that not enough credence was given to clandestine reporting about the strategic missiles, reporting which in his view pointed accurately both to the decision to deploy them and to the actual deployment. We will answer this first.
The Clandestine Reporting
In the course of one of our two studies, the vast body of this reporting -- comprising more than 14,000 reports -- was examined minutely. The reports cited by Mr. Rumpelmayer were included in this review. It was our judgment that the bulk of this material could not have been evaluated with confidence at the time unless information were available from other sources against which it could be checked, and that there would have been no way to identify the "rare nuggets" among the "tons of dross." We found, indeed, that even in retrospect one could not construct from the clandestine reporting alone a coherent account of the course of the venture.
With respect to Soviet intentions, of the six reports disseminated before 19 September -- the date of the estimate -- which are said to point to those intentions, the first and sixth do point vaguely in that direction. But the first seems to be canceled by the second (which makes both look like reports of Cuban hopes, not Soviet intentions, and moreover of hopes which would have lagged behind the actual agreement by about three months); and the sixth does not specify the kind of "rocket project" (SAMs were known to be on the island). Even if the source had been known to be reliable (and he was not so known), neither report would seem to take the analyst as far as Raul Castro's public boast, in the same period, that his negotiations with the Russians had changed the balance of power in the world -- a remark which precisely described the aim of the missile base venture.
Similarly with respect to the actual deployment, of the ten reports cited by Mr. Rumpelmayer under "signs of deployment" not one carried a source appraisal that would have commanded the credence of an analyst at the time. Three were evaluated F-3 and two F-6; five carried no evaluation. The first four of the six disseminated before 19 September would not be accepted even now as accurately reporting on deployment (although the third, dated 31 August, described credibly preparations for receiving missiles), because there is pretty good evidence that no strategic missiles arrived before mid September.3
And the later reports (in addition to the good report of 31 August) would have had to struggle for acceptance, because all such reports had to be read in the light of the many false reports of previous months: there had been more than 200 reports of Soviet missiles in Cuba before January 1962, and the many reports of construction activity and equipment observed during the spring of 1962 (some in areas where SAMs were later discovered) had been negated by photography of those areas during or after the reported period of observation.
It was never a question of "majority rule" but rather, as always, a question of credentials, grounds for credence. Majority rule, with no attempt to discriminate, would have produced an estimate favoring the large number of affirmative reports (long before the fact) over the smaller number of negative photographs. But the credentials of the two sets were very different. As for discriminating among the reports themselves, Mr. Rumpelmayer makes clear, with respect to the single good source of all of the reports cited as pointing to Soviet intentions, that it was only afterwards that he was checked out as a reliable source; in other words, his credentials had not been established even to the collector's satisfaction at the time. Was there an) source with established credentials who was ignored?
As a matter of fact, these reports that could not be accepted without corroboration were not ignored, were not "set aside." They had long provided a part of the reason for conducting systematic aerial reconnaissance of Cuba. And during September, when the reconnaissance flights were basically peripheral and did not provide thorough coverage of the island, these reports did the best thing they could have done: they set off the process which led to the collection of photographic evidence on 14 October. That is, by late September those making the decisions had concluded -- mainly on the basis of the sharper and more credible reports after mid-September -- that whatever the reasons for contenting ourselves with less than thorough coverage, there were better reasons for making a maximum effort; and this decision was vindicated by the first subsequent flight.
The Soviet Attitude
Mr. Rumpelmayer puts his other contention -- about Soviet estimates -- in these terms: the Soviet leaders were willing to increase greatly the level of risk because the gains to be made from a successful venture were so great-but they were prepared to withdraw "if caught." To take the second part of the proposition first, this is manifestly false unless one assumes that the missiles were to be used solely for a surprise attack on the United States; their usefulness for anything else depended directly on their being "caught," on their presence being known when the program was completed.
Suppose we replace this second part with a formulation that many observers would accept-that the Russians were prepared from the start to withdraw (as Mr. Rumpelmayer says later of the actual disengagement) as soon as "convinced that the United States was ready to act" But if we define the risk as the risk of U.S. military action against Cuba or the USSR, then the first part of Mr. Rumpelmayer's proposition is also false. That is, if Khrushchev was confident that he would be given time to withdraw-that the venture would be accepted as a "mere probing action" -- then he was not consciously accepting a high risk of this kind.
But what of the other kind of risk, the risk of failure, of a humiliating withdrawal in the face of an American ultimatum? As Mr. Rumpelmayer puts it, the Russians would not have had ready a "contingency plan" if they had really been wrong about the "probable U.S. reaction."
Taking these words at face value, Khrushchev regarded U.S. willingness to fight as the probable response, thus expected the venture to fail, in other words was one of those sports who fly into Las Vegas hoping for the best but cheerfully expecting to go home without a dime (or with just a dime). This proposition is not patently false; but the evidence does not support it.
Our two studies, cited above, agree Khrushchev recognized from the start that there was some degree of risk of an American attack at one or another point in the venture but believed this risk to be small. As witness, the Russians were aware of U.S. reconnaissance capabilities but did not camouflage the strategic missiles or conceal their deployment, left the MRBM sites identifiable for a long period prior to the establishment of an IRBM capability (which would have completed the program), and did not employ their air defense system against U.S. reconnaissance aircraft. Moreover, they sent the missiles into Cuba after President Kennedy's firm and explicit warnings of early September, without knowing that the pattern of reconnaissance was to be changed to their advantage.
Similarly, the two studies agree that Khrushchev recognized from the start a possibility of failure but believed at least until September -- perhaps until mid-October -- that the United States would probably acquiesce and until late October that the venture could be managed to his profit even if the United States did not acquiesce. He seems to have calculated -- judging from some of his statements before the crisis and his conduct during the week of the acknowledged crisis -- that the United States would at most impose a blockade and could probably be tied up in negotiations in the course of which he could either complete the program (and thus increase his deterrent) or at least gain large concessions for withdrawal.
We agree with Mr. Rumpelmayer that Khrushchev had so much of a "contingency plan" for withdrawal. But the character of the "plan" is one of our reasons for thinking that he did not expect to fail. Much of his behavior in the week of the crisis seemed improvised and erratic: he lied about the missiles after their presence had been established beyond doubt; he continued work on the bases while frantically attempting to pacify the United States; he threatened to run the blockade after ordering his ships to turn around; he warned that he would fire the missiles and at the same time promised not to; he transmitted an explicit offer of withdrawal for a no-invasion pledge before transmitting his letter implying a willingness to withdraw; he made his Cuba-for-Turkey proposal after having implicitly and explicitly offered a better deal; and he finally accepted the proposal which President Kennedy attributed to him, a capitulation at Castro's expense, without consulting Castro. This was hardly the smooth performance of one who had been expecting to be forced to withdraw.
Thus we agree with Mr. Kent that Khrushchev made a serious mistake in judgment. He seems not to have recognized what American estimators recognized and, not unreasonably, expected him to recognize -- that, if Soviet gains from a successful venture were to be so great, it was probable that the United States would recognize what was at stake and would act to deny such gains to its principal antagonist, just as President Kennedy had repeatedly told him.4
The Real Lessons
We too have some opinions as to lessons to be learned from the Cuban venture -- lessons which if learned would reduce the possibility of "misestimating the Soviets in a future case."
The lesson for the collector is obvious: that he cannot expect his good reports to be recognized and accepted at once if the record of reports from the same kind of sources is a poor one. ("Wolf! Wolf!" )
The lesson for the estimator might be this: to allow a bit more, regularly, for Khrushchev's disposition to wishful thinking and for his inclination to commit himself to a serious action without thinking it through. He seems to have just now (August) done it again, in making and publicizing his arrangements for the meetings of the Communist parties.
For the analyst, the lesson might be to give Soviet public and private statements the closest possible scrutiny, looking at them again and again until satisfied that all of the possible implications have been recognized. For example, the Soviet insistence on the formula of the defensive purpose of the weapons in Cuba seems to have been in part a means of inviting the United States to acquiesce in the deployment under this euphemism. Similarly, some Soviet statements seem to have been implying an offer by Moscow (another observer has privately made a good case for this) to continue doing what it was doing in Cuba but not to embarrass the Administration by revealing this until after the U.S. elections -- at which time the program would be complete -- if the United States did not reveal it. A sense of these implications need not have led to the conclusion that the Russians did indeed intend to deploy strategic missiles in Cuba, but they might have placed stronger qualifications on the official judgments of what Moscow was up to and might have led to earlier warning.
There may be lessons for the policy-maker too. One of these was apparently learned very rapidly and expressed in the decision of late September to restore the pattern of thorough aerial reconnaissance over Cuba. The lesson was that a nation might be embarrassed by the utilization of a given intelligence asset but might be destroyed by the failure to use it.
The other lesson relates to the adversary's reading of U.S. behavior. As the observer cited above was the first to remark, no American official (so far as we know) chose to question the Russians directly about this crafty formula of defensive purpose, to ask a knowledgeable Russian official point-blank just what kind of weapons were going into Cuba -- a restraint which may have encouraged Moscow to believe that its invitation was being accepted. And there were other features of American behavior, including the failure to make a maximum reconnaissance effort between 5 September and 14 October, which could have been read by Moscow as indicating tacit agreement. The policy-maker may be able to use more help than he normally gets in judging how the signals he is sending -- or things that he is doing which may be taken as signals -- will be read.
4 Khrushchev's failure to recognize this was foreseen by the only official known to us to have predicted the deployment of the strategic missiles in Cuba. Challenging the draft estimate, this observer argued that Khrushchev might well be so dazzled by his possible gains as to be unable to recognize the true level of the risks.