Whatever Nixon’s view of the Agency, it was my opinion that he was the best prepared to be President of any of those under whom I served—Eisenhower, Kennedy, and Johnson. Each had great strengths, but as I saw it, Nixon had the best grasp of foreign affairs and domestic politics. His years as vice president had served him well.
As Henry continued to talk, I realized that, in effect, Nixon had reverted to an expanded semblance of the National Security Council that President Eisenhower had established. Neither President Kennedy nor President Johnson was much interested in the NSC, and made little use of it. Under Nixon, and with Henry Kissinger’s leadership, the NSC assumed a much more significant role in policymaking. The personnel strength quadrupled in size and grew in substance, with the addition of a number of excellent analysts detailed to the White House from the Agency and State. Various NSC subcommittees were formed with experts also borrowed from the Agency, State, and Defense. The most important of these bodies was the Senior Review Group, which studied policy recommendations before they were presented to the full NSC. Jack Smith was an obvious choice to represent the Agency as a member of this senior group.
In the course of attempting to shape our product to the requirements of the new administration, Jack’s initial conviction seemed confirmed. Nixon was not reading the PDB—still the most important daily intelligence document in Washington. I agreed with Jack’s suggestion that he raise the topic with Kissinger, with whom we were developing a working relationship. Jack arranged a meeting, and by luck found Attorney General John Mitchell in Henry’s office. When Jack raised the question of the PDB, Mitchell got up to leave. Henry requested him to stay, and asked if he knew what Nixon thought of the PDB. Mitchell said he thought that Nixon considered the document too far-ranging, and that it covered areas which were of slight interest to the White House. He added that Nixon had once said he had trouble sorting fact from opinion in the PDB.
Jack then asked if Nixon wanted just the facts and no interpretation. Mitchell, who was less than expansive at the best of times, was downright laconic with relative strangers. He grudgingly allowed that Nixon was a lawyer, and lawyers always liked to have the facts first, with the opinions to follow. It was not much to go on, but Jack began refashioning our efforts accordingly.
Nixon’s decision that I was to decamp before the National Security Council began discussing policy matters struck me as shortsighted, and not merely because of any anguish the group might suffer from my absence. Minus the director of Central Intelligence, there would be no one left standing who on occasion might be able to “keep the game honest” by pointing out that some of the bona fide NSC members were making too freely with dubious data.
At the next NSC session, I was preparing to slip gracefully away when Nixon invited the group, one and all, for lunch with him. In the NSC meetings that followed, I thought that Nixon had either forgotten his initial ukase or had clearly changed his mind. I was thereafter allowed to remain at table as the NSC went on about its policymaking business. Sometime later I was told that Melvin Laird, secretary of defense, had interceded with Nixon, and insisted that the DCI be present at policy discussions.
Very early in my career I realized that secret intelligence is not for the fainthearted. From the mortal peril of organizing resistance and stealing secrets in police states to dealing with one’s own government, secret intelligence can be a lethal version of a rugged contact sport. In the Johnson administration, the near nightmare of balancing the conflicting Pentagon and Agency views on the strength of the North Vietnamese forces in South Vietnam had been resolved. But blood was on the floor and walls before a compromise that did not violate the integrity of the estimate had been agreed upon, and President Johnson was provided with the most accurate possible estimate of the North Vietnamese forces. In September, an equally explosive issue came to the flash point in the Nixon administration.
A New England-born colleague characterized the battle between the Agency and the Nixon Pentagon over the capabilities of the newest Soviet intercontinental ballistic missile (ICBM) as a “real pea-whistler.” He explained that in Maine there were two varieties of pennywhistles available at every corner candy store: one cost a nickel, the other a dime. The price difference was mandated by the presence of a dried pea in the luxury whistle. For some reason the pea raised the volume of the classy whistle to an excruciating level. Hence, any top-of-the-line product was referred to as a “pea-whistler.”
The Agency had designated the USSR’s new ICBM as the SS-9. It was the first Soviet MIRV (multiple independent re-entry vehicle). A MIRV is a rocket that carries more than one warhead, each of which separates from the mother rocket before striking individual targets. Our data on the SS-9 were too scant to permit a firm estimate of its capability. The first issue was the throw weight. How much of a load could the MIRV carry? The Air Force, impressed by the SS-9’s bulk, granted it an enormous capability. Our experts thought it less powerful. The Pentagon military analysts considered the weapon to be highly accurate. The Agency team disagreed. Although all the disputants agreed that the SS-9 was built to transport three independent atomic warheads, there was more to come.
In mid-1969 the Nixon administration was pushing for congressional approval of a multi-billion-dollar program to develop an anti-ballistic missile (ABM). On form, and if such a defense weapon could be perfected, the ABM program would provide rockets which could detect and destroy incoming ballistic missiles. Critics of the ABM program described it as a bullet intended to shoot down another bullet. The Pentagon, with its eye on convincing Congress that an American anti-ballistic missile was an essential defense need, decided that each of the three SS-9 atomic warheads had its own independent guidance system that would steer it to pre-selected targets. If anything was likely to unleash the dollars needed to create an ABM, the specter of a score of SS-9s delivering sixty precisely guided missiles in one volley should have carried the day.
Agency analysts disagreed, and remained convinced that any such independent guidance capability was beyond the grasp of Soviet science, and the research and testing so expensive it might unhinge the USSR’s economy. As the estimates’ battle raged—and this is not too strong a word—the Pentagon retreated a bit. Perhaps the three warheads would not have independent guidance systems. Even so, twenty SS-9s, each loaded with three powerful atomic warheads—to be flung like fistfuls of golf balls in a predictable pattern that would cover most of the chosen targets—could be presumed to destroy all the Pentagon silos enclosing the Minute Man rockets. CIA still disagreed: the Russians did not have and were not likely soon to achieve even that level of guidance capability.
As Jack Smith has pointed out,* this led to another even more vital conflict. Was the USSR striving to achieve a first-strike capability? In military terms, “first strike” means much more than landing the first blow. A successful first-strike ICBM salvo will demolish the enemy’s ability to return fire. By definition, a first strike is a knockout blow.
The Agency position was firm. The USSR was not seeking a first-strike capability, and the SS-9 was some four years away from its first testing.
Secretary of Defense Melvin Laird was not persuaded by the Agency view of the SS-9 capability. Having been argued to a standstill by CIA, the Pentagon turned to TV. Displaying a variety of charts, Pentagon spokesmen illustrated the SS-9 “footprint,” a homely military term for the predictable pattern in which the individual warheads would presumably return to earth. This, it was explained, would be accurate enough to demolish the U.S. missile silos. These briefings and press leaks, some of which charged the Agency with a bias against the Pentagon position, raised the debate temperature to a boil. At one point, I was moved to instruct Agency officers that under no circumstances were they to make any public statements—pro or con—on the issue.
In early September the Agency submitted the annual updating of the earlier Top Secret National Estimate of Soviet ICBM capability (“Soviet Strategic Attack Forces”). A brief paragraph
in the updating reaffirmed the earlier estimate—the USSR was not seeking a first-strike capability. It also stated that the Soviets had not yet produced an SS-9 for testing, and expressed doubt that this could happen before 1974. As always, this document was presented to all of the relevant U.S. intelligence and policy offices for coordination. It was produced with many dissenting opinions cited in footnotes. The Nixon administration reacted at once. Mel Laird was about to give a speech outlining the administration’s policy on first strikes and MIRVs. Where, he demanded, did CIA get off contradicting Nixon’s policy?
When Laird’s complaint hit my desk, I summoned the Agency experts. Not one of our analysts or weapons specialists agreed with the Defense Department position. The resulting standoff was as difficult a policy problem as I had faced. I realized that there was no convincing evidence in the Agency or at the Pentagon which would prove either position. Both positions were estimates—speculation—based on identical fragments of data. My decision to remove the contested paragraph was based on the fact that the Agency’s estimate—that the USSR was not attempting to create a first-strike capability—as originally stated in the earlier detailed National Estimate would remain the Agency position. The issue came down to whether or not the initial estimate needed to be stressed in this updating of the original document. In the interest of allowing the Defense Department analysts and weapons experts to state their position, I agreed to drop the disputed paragraph.*
An important factor in my decision to compromise with the Defense Department position was my long-standing conviction that it is a serious mistake for any intelligence service ever to assume that it has achieved absolute wisdom. I disagreed with the Pentagon position, and could not suppress the feeling it was tainted by the Nixon administration’s determination to develop an anti-ballistic missile. Although the Agency’s initial estimate would stand, I concluded that the Defense Department analysts and weapons experts should have ample opportunity to state their position.
At the height of the debate the atmosphere was further poisoned by a remark attributed to senior members of the administration: “Whose team is CIA on?” In other words, “Let’s all get together and trim the evidence to suit the wishes of the politicians.”
Aside from Jack Smith, whose view of the situation coincided with mine, the decision did not sit well with the Agency analysts and experts. In their view, I had compromised one of the Agency’s fundamental responsibilities—the mandate to evaluate all available data and express conclusions irrespective of U.S. policies. Pertinent to this discussion is the fact that the President is under no obligation to accept any intelligence estimate at face value, nor is he required in any way to follow it. The President’s responsibility is to act in the best interest of the country. I was not prepared to stake the Agency’s entire position on this one issue—in an average year CIA was making some sixty estimates, very few of which ever reached the President’s level of concern. I was convinced we would have lost the argument with the Nixon administration, and that in the process the Agency would have been permanently damaged.
The final step in the MIRV conflict was taken by Henry Kissinger when he informed me that the National Security staff would examine all the data and come to its own conclusion. Jack Smith and Abbott Smith, head of the Office of National Estimates, conferred with Kissinger. Henry was clearly in no mood to accept the Agency position as stated in the study. Jack’s offer to rephrase the data that Henry had considered biased was refused. The NSC staff would make its own evaluation. Several days passed before we learned that this staff had made no changes in the CIA estimate of the SS-9’s performance.
The Agency’s estimate that the USSR would not have a MIRV/SS-9 first-strike capability for some four years was borne out in 1974 when the Soviets tested their first MIRV. The timing of this test tallied nicely with the Agency’s original estimate.
As the dust settled, the Agency encountered another collision with the Nixon administration.
*When Jim Angleton’s full name became known within the Agency, a friend noted that a powerful name was obviously one of the criteria for senior operations officers—James Jesus Angleton, William King Harvey, and Thomas Hercules Karamessines.
*Russell Jack Smith, The Unknown CIA (Washington, D.C.: Pergamon-Brassey’s, 1989), p. 207. This is an excellent text on the Directorate for Intelligence, and a fundamental document in CIA history.
*Ironically, Thomas Hughes, head of the State Department intelligence office, reinstated the dread paragraph as a footnote, regretting its omission in the original text.
Chapter 37
—
SIHANOUKVILLE
The strenuous tussle over MIRV capabilities and the SS-9 coincided with, but did not interrupt, the Agency’s battle to produce the necessary intelligence on North Vietnam. We were trying every operational skill in our arsenal to penetrate the policy levels of the government of North Vietnam. I kept a high standard of staff personnel in Vietnam, and the best of the Agency operatives around the globe were focused on Hanoi. Nothing worked. And there was no satisfaction in the knowledge that we had done our best. In the years since President Nixon closed the door on Vietnam, I have yet to uncover any means that we did not exhaust in the effort to reach Hanoi.
Strategically, one textbook approach to countries as alien to our culture as Vietnam might require a ten-year commitment. It would take that much time to develop a sophisticated cadre of language and area experts prepared to make their career in such remote areas. In a decade, the right operatives might achieve command of the necessary languages, and the desired in-country familiarity with the cultures and terrain. With commitment and effort, there might develop personal relationships—and identity of interests—strong enough to survive political fallings-out that might follow. The possibility of funding such an effort in half a dozen likely areas and finding personnel willing to make the necessary commitment to the job would be a major undertaking. In the absence of any such unrealistic strategic commitment, the tactical solution is obvious. Do not intervene on the ground unless prepared to make an all-out effort. And do not count on a secret intelligence agency being able to pull political folly away from the blaze.
Given a choice, I would have picked a less dramatic proof of my conviction that intelligence services have never come close to possessing divine insight. But no such luck, Sihanoukville it was to be. For years the Agency and the various military intelligence services had been monitoring the North Vietnamese incursions into Cambodia and attempting to determine the amount of Chinese military matériel coming through the Cambodian port at Sihanoukville en route to the North Vietnamese forces.
There was no doubt that North Vietnamese troops and Viet Cong guerrillas slipping down the Ho Chi Minh Trail would dodge across the border into Cambodia to regroup, refit, and catch their breath before continuing. The open questions: how many troops are in transit, and how much matériel are they toting? The Agency estimate was that some four or five times more matériel was coming down the trail than before ROLLING THUNDER, the Air Force’s widely publicized attempt to interdict the traffic. This antagonized the military staff in Vietnam (MACV) and the Pentagon, and can only have dismayed President Johnson.
In contrast to the CIA’s glum estimate of the traffic on the Ho Chi Minh highway, our estimate of the Chinese-supplied matériel transiting Sihanoukville was almost exactly half the tonnage that MACV and the Pentagon analysts had determined. Both the Pentagon and Agency teams worked from the same data—scraps of hearsay, lowest-level agent reports, and unreliable prisoner-of-war interrogations. Neither MACV nor the Pentagon analysts saw reason to give ground to the equally entrenched Agency experts. In almost every respect this was a sour replay of the “First Strike/SS-9/MIRV” controversy: competent experts in honest disagreement.
As with everything in Vietnam, there was intense pressure from all sides. At a morning staff meeting, I asked, in the absence of any new information on Sihanoukville, how we might at least resolve the c
ontroversy over the existing data. Jack Smith’s suggestion that he go to Saigon and talk with the MACV experts was immediately agreed upon. He would determine how the MACV staff arrived at its estimate, and learn if, by any good fortune, the military had come upon an informed source that CIA did not know about.
Jack was impressed by the quality of the MACV analysts and their straightforward presentation of their position. The data were almost identical with the Agency’s. And as Jack reported, “Alas, it was the same shoddy, low-grade information we were using.” The MACV deduction remained firm: twice as much matériel was flowing through Sihanoukville as CIA estimated. In May 1970, Secretary Laird and General Wheeler testified before Congress using the figures that contradicted the Agency estimate.
Dawn broke in late June, when CIA operatives recruited an agent with direct access to the bills of lading that provided documentary evidence on the shipping passing through Sihanoukville. It was a classic bit of espionage, involving spies, secret writing, radio communication, special cameras, and film. The result was a steady flow of documentary intelligence on the Chinese arms shipments moving through Sihanoukville to staging areas in Kompong Speu and eventually to the “Parrot’s Beak” and Vietnam. Photographs of the bills of lading were supplemented with local agent sightings of the after-dark off-loading of the equipment. At one point, CIA bought a truck which was accepted into the convoys that were moving the arms shipments.
The carefully recorded data showed conclusively that the Sihanoukville matériel was much more than the CIA estimate, and, if anything, a bit more than the MACV-Pentagon figures. In one outstanding espionage operation, we managed to prove that the CIA estimate was more than 50 percent off base.
A Look Over My Shoulder Page 47