Lessons To Be Learned About the Intelligence/Policymaking Relationship ⁽¹⁾

The mixed picture of CIA performance illustrated in these episodes should not obscure the generally good analytic record the Agency chalked up on Vietnam in the years under review. From the early 1950s onward, CIA's assessments in the main proved more accurate than those of any other US Government entity, and CIA's analytic record on Vietnam compares favorably with its endeavors in the counterinsurgency field. CIA officers fairly consistently insisted their analyses showed that military force alone would not win the war; that our South Vietnamese creation, the GVN, was not proving adequate to the political-military task; that we should not underestimate the enemy's covert presence throughout South Vietnamese society; that we should not underestimate the enemy's staying power; that US bombing efforts were not appreciably slowing the enemy's progress in the South; that the enemy would try to match US escalation rather than meaningfully negotiate; and that ill-founded official claims of great progress distorted reality to the detriment of policy objectives.(2) CIA's record of candor is all the more remarkable because CIA officers often had to brave pressures from senior political and military officers to "get on the team" and to support the war effort with more optimistic findings and estimates.

That overall record must be tempered, however, by the fact that on the three occasions under review in which Agency assessments had a chance to affect key US decision points in Vietnam--how to assess and deal with a failing Diem regime, whether and how to "go big" in Vietnam, and how to assess the enemy's subsequent capabilities and intentions--the character of CIA's intelligence input was mixed.

In our first episode, 1962-63, because DCI McCone brought heavy pressure on the Board of National Estimates and the Intelligence Community to produce a more optimistic National Intelligence Estimate than they felt the evidence supported, because the Board caved in to that pressure, and because that NIE fed the confidence of policymakers that the war effort was going fairly well, those policy managers were wholly unprepared to deal with the sudden collapse of political stability in South Vietnam only days after the definitive NIE was issued. The judgment must be made, as we have seen in the statements of the authors of The Pentagon Papers, that in the case of that NIE senior decisionmakers were influenced by an Intelligence Community product; the problem was that, thanks to the intervention of policymakers and program managers, this particular input was misleading.

Director McCone's subsequent cautions concerning the wisdom of overthrowing Diem proved well taken, but they failed to counter the original impetus or the momentum that gathered around that impulse. Equally unfortunate, the sorry outcome of the unilateral Harriman-Hilsman-Lodge initiative had the ironic result of leaving CIA with much of the blame for the disaster, obscuring the fact that the CIA Director had tried his best to persuade the White House that that course would breed disaster.

In no period during the Vietnam conflict did the conclusions of CIA's working-level officers prove more accurate in retrospect than in the second episode we have examined, when they consistently argued in 1963-65 that substantially increasing US combat operations in Vietnam would not solve US problems there because the war was essentially a political-military struggle which had to be won in the South and primarily by the South Vietnamese. Those arguments unfortunately made little if any dent in policymakers' increasing certainty that the war in the South could be won only by committing US forces to combat there and by consistently bombing the North. One of the reasons the impact of those assessments was dulled was because the Agency spoke with two voices at the time, with Director McCone giving the President's circle his personal assessments which at times did not agree with those of his analysts and officers. But as we have seen, other, more potent forces also caused policymakers to shrug off CIA assessments they found uncongenial. In only a slow, cumulative sense did the Agency's generally pessimistic analyses find resonance among some senior consumers and contribute to the growing uncertainty of Robert McNamara, George Ball, Hubert Humphrey, Clark Clifford, and the civilian heads of the Pentagon's International Security Affairs bureau that there was light at the end of the tunnel.

In our third episode, 1967-68, a few working-level CIA officers developed and championed accurate assessments that enemy strength in South Vietnam was perhaps twice what the US military was willing to acknowledge, and that the enemy was about to change strategy radically by launching a nationwide offensive. Many hazards, however, undercut those judgments. Political pressure from the White House, MACV, and the US Embassy to understate the number of enemy forces caused DCI Helms, Special Assistant George Carver, and the Board of National Estimates to override the conclusions their analysts had derived from available evidence. Then Headquarters analysts themselves refused to accept new field estimates of the enemy's intentions for Tet because these did not jibe with their own published estimation of the enemy's likely conduct.

Last but not least, CIA's offerings to senior managers of Vietnam policy generally had to filter through one particularly influential Agency officer, George Carver, who until two months after the Tet Offensive generally supported the Johnson administration's view that things in Vietnam were looking up. To his credit, in the end Carver did level with the President and his "Wise Men"--and so helped influence the radical changes Lyndon Johnson began to make in his domestic and Vietnam policies at the end of March 1968.(3)

In our three episodes, why were CIA's published analytical judgments so often more pessimistic than the positions of the rest of the government? By and large, CIA's analysts had no special sources of intelligence not available to others; the difference was in the interpretations they gave existing evidence. In this they had certain advantages over other US Government analysts. They were much freer at that time from pressures to produce judgments supporting the operational offices' enthusiasms that the tide in Vietnam was being turned by the efforts of the US Mission and MACV. Also, CIA reporting from the field was generally more rigorously conducted and more candidly transmitted than that of most other USG elements, and it gave Headquarters analysts unique insights into political developments in South Vietnam. Compared with their colleagues in the military, DDI and O/NE officers had usually been at their jobs longer and were more experienced at interpreting and calling developments in Indochina. And many of CIA's Vietnam analysts of this period were the recipients of occasional confidences from working-level field officers, civilian and military, about the difficulties and distortions they were encountering in their attempts to get candid reporting past their superiors in Saigon. By 1966, improved CIA field reporting--especially overhead imagery--and new analytical methods enabled CIA officers to quantify North Vietnam's continuing ability to support its forces in the South despite the US bombing campaign. It was these methods and judgments that at last helped convince the Defense Intelligence Agency and Defense Secretaries Robert McNamara and Clark Clifford that America's expanded efforts were not causing the enemy to slack off, and probably would not do so.(4)

The advantages enjoyed by CIA officers deserved greater respect from decisionmakers than they received. And the credentials CIA's analysts brought to bear on Vietnam issues were impressive. Many of these officers had not only been studying East Asia questions for years, but also had racked up fairly strong estimative batting averages. They had correctly warned that the difficulties the French were encountering in combating the Communist-led Viet Minh's military and political advantages would force Paris to call it quits in Indochina--and so confront the United States with a very weak South Vietnam and very difficult policy decisions on whether and how to take up the anti-Communist burden there. CIA officers had accurately gauged the limits of Communist China's boldness in the Quemoy-Matsu offshore island crisis of 1958. DDI and O/NE officers, moreover, had led the way within the Intelligence Community in trying to alert decisionmakers that Moscow and Beijing would split--a development of immense consequence that had become clearly evident by 1963-65, the key years in which President Johnson and his advisers were wrestling with the question of whether to go big in Vietnam.

Questionable on several scores is former Secretary of Defense Robert McNamara's present complaint that "there were no Vietnam experts" to whom policymakers could turn for advice. That charge not only reveals that he and his colleagues were ignorant of the credentials which Agency and other Intelligence Community officers brought to their tasks; it also contradicts the several acknowledgements McNamara himself makes in his In Retrospect to the positive contributions DCI Helms and other CIA officers gave him.

During the years under review here, the dangers of offering heretical expert counsel were repeatedly demonstrated. Senior policymaking officers often put pressure on analysts to make their assessments rosier, or placed political lids on objective estimates of the enemy's strengths, or warned doubters to shape up and join the team. Especially illustrative in these respects was the fate of State Department expert Paul Kattenberg who, after telling President Johnson and other senior officers that available evidence did not support the optimism they were expressing, was relieved of his Vietnam responsibilities and thereafter given backwater assignments.

Thus, although there were ranks of competent CIA Vietnam experts ranged behind John McCone, Richard Helms, George Carver, Bill Colby, and the few others who dealt with top Administration officials, they were perceived as juniors of unknown quality by most senior decisionmakers. Perhaps contributing to this perception was the fact that, at least until 1967-68, CIA's views on Vietnam were generally uncongenial to most policy planners. Not only did many CIA officers question the progress so many top officials were claiming, they also did not accept the widespread assumption that Vietnamese Communist aggression was essentially one thrust of a global campaign of conquest masterminded in Moscow and Beijing. Not least, O/NE's officers had the audacity to doubt the core belief of the American political-military establishment that the fall of Saigon would necessarily lead to an inexorable Communist takeover of all Southeast Asia. Of such elements was the judgment made that there were "no experts at hand."

It must nevertheless be emphasized that the analysts of CIA's O/NE, OCI, and the Office of Research and Reports did not fully agree among themselves on all questions about Vietnam. Though they shared a generally more pessimistic view of events there than did most Vietnam analysts elsewhere in the government, there were differences on various questions among offices and analysts. The least pessimistic tended to be officers of the North Vietnam Branch of the Office of Current Intelligence, who were by definition analysts of global Communism and had served in Soviet-related offices.

In any case, among the principal intelligence concerns these episodes illustrate is the mischief that can be done by stubbornly held preconceptions, and by the unwillingness of senior consumers of intelligence to entertain new data or judgments that do not support their own analyses or that threaten their political commitments. Such hazards clearly were important factors in the refusal of policymakers to buy the working-level intelligence officers' warnings in early 1963 that all was not well in South Vietnam; and later, in CIA's insistence that going big in Vietnam would not do the trick; and still later, in Saigon Station's warning that the enemy was about to launch a major nationwide offensive.

These episodes also show that preconceptions are not the monopoly of policymakers. Intelligence analysts and managers, too, can be unreceptive to new, different stimuli. This certainly occurred in early 1963, when John McCone refused to credit O/NE views which challenged policymakers' optimism; and in the winter of 1967-68, when George Carver and CIA Headquarters analysts preferred their existing estimates of likely enemy behavior to Saigon Station's new and different interpretations.

We have also seen that preconceptions were at times more firmly held and resistant as one went up the lines of command in CIA, in somewhat similar fashion to the hesitance of senior military and Embassy officers in the field to accept their junior officers' more candid reporting and assessments.

Then, too, these episodes show also that intelligence facts ultimately are no match for political considerations. Such concerns distorted intelligence in early 1963. They accounted later for the deaf ear that policy managers turned to CIA officers' skepticism about "going big" as a cure-all in Vietnam. Political considerations distorted intelligence again in 1967 when fear of political embarrassment dictated that only so many enemy troops could be counted and that the Viet Cong's additional, irregular forces could not be counted at all.

CIA officers who serve at the interface of intelligence and policy are no less subject to the inherent conflicts between the two, and when a policy problem lasts as long as the Vietnam War did, the infection of intelligence estimates by policy concerns is inevitable. At crucial points in the three episodes studied, some senior CIA officers felt they had to adjust what might be called "pure" intelligence judgments to "practical" political considerations, as did the Board of National Estimates in its 1963 NIE on the outlook for Vietnam; DCI Helms in the 1967 faceoff with MACV's Order-of-Battle estimators; O/NE and the Intelligence Community in the subsequent NIE on the enemy's O/B and combat capabilities; and Vietnam Special Assistant Carver in both the O/B controversy and the Tet offensive forecasts.

These episodes also illustrate that one reason CIA's inputs did not have greater impact is that the Agency's officers often defined their intelligence roles rather narrowly. Policymaking customers have often complained that intelligence officers, by being too shy about intruding into policy matters, fail to offer up the helpful policy-relevant ideas of which they are capable. There is indeed a fine line that intelligence officers must follow in order to retain their essential credibility as policy-free advisers.

Yet it seems clear that when CIA officers did volunteer policy critiques in 1967-68, their arguments definitely contributed to the diminishing certainty among Administration officials--including Secretaries McNamara and Clifford, and belatedly President Johnson himself--that a military solution in Vietnam was possible. One of the most remarkable of these, one that went considerably beyond the strict lines of intelligence matters, was a sensitive assessment sent "Eyes Only" to the President by DCI Helms well before Tet, in September 1967. As cited by Robert McNamara in his retrospective book, that memorandum concluded that, although an unfavorable outcome in Vietnam would of course have many very damaging effects, "The risks are probably more limited and controllable than most previous argument has indicated."(5) Observing that the assessment showed that "CIA's most senior analysts believed we could have withdrawn from Vietnam without any permanent damage to U.S. or Western security," McNamara states that the CIA authors were expressing the same view he himself was giving to Senator John Stennis's subcommittee at the time, "supported by CIA/DIA analyses," that "we could not win the war by bombing the North."(6)

Similarly, certain of the candid policy prescriptions pushed forward by CIA officers in February and March 1968 doubtless added to the post-Tet recognition by policymakers that US policies sorely needed re-examination. As we have seen, such CIA assessments certainly included Bill Colby's "Operation Shock" recommendation in February 1968: that if the GVN did not shape up, President Thieu should be advised that the United States would reserve its position with respect to South Vietnam. Even more frank were the views of CIA analysts given to President Johnson by DCI Helms that same month: that the United States had no strategic or coherent policy in Vietnam, and that if the GVN proved unable to make its way, the United States ought to get out. And we have seen also that George Carver was in constant contact with Lyndon Johnson's closest advisers throughout February-March 1968, and that in the end he gave them candid appraisals of the limited prospects in Vietnam.

Another facet of intelligence-policy interplay illustrated in these episodes is the key influence that individuals often exert on either side of the equation. John McCone was much less hesitant to offer up policy-relevant inputs than were DCIs before and after him. George Carver, though a middle-level officer, enjoyed extraordinary entree and influence with top policymakers. Assertive, strategically placed officers in the policy hierarchy, notably Robert McNamara, Walt Rostow, McGeorge Bundy, Roger Hilsman, and Robert Komer, often overrode competing arguments. And Lyndon Johnson's hardly subtle influence was paramount.

We have seen, too, that the views individual officers hold may differ from time to time, depending on what responsibilities they hold and whether they are speaking privately or for the record. Among intelligence officers this was shown especially in the degrees of candor displayed by George Carver when he was an O/NE officer remote from the policy arena and when he was the DCI's Special Assistant for Vietnam Affairs. Policy officers' statements (illustrative examples are in the Appendix) could be influenced by the office they happened to hold at the time, and by whether they were giving higher authority their private views or publicly conforming to the official line of the moment. Once they were out of office, moreover, hindsight could alter an earlier view.

Salient in these episodes is the fact that the intelligence and policy worlds often were widely separated, their respective officers ignorant of the other's world. If policymakers were at times unappreciative of how Intelligence Community experts could help them, it was equally true that intelligence officers might have carried more weight if they had been closer to the decisionmakers. Most of the CIA officers who earnestly offered up their assessments on Vietnam were separated by a figurative and literal river from the policymaking arena. And rarely, if at all, did they factor in the many broader questions with which their seniors had to wrestle or consider the many inputs other than intelligence which of necessity influence the determination of policy. The intelligence officers' task studied in our episodes was a narrowly focused one, and their vision was largely confined to Vietnam and Southeast Asia, whereas policymakers saw the Vietnam problem as one of many in much broader perspectives, both foreign and domestic.

Indeed, these episodes demonstrate that the impact of intelligence on policymaking consumers is clearly secondary to that of broader, outside forces. CIA's pessimism regarding proposals for sending US forces directly against Hanoi and its expeditionary forces was offset by Lyndon Johnson's determination to succeed in Vietnam. It seems clear that John McCone almost instinctively deep-sixed the challenged draft of NIE 53-63 because his confidence in his Board of National Estimates had been shaken by the Board's earlier miscalling of the emplacement of Soviet missiles in Cuba, and because he was still under fire on that issue from the President's Foreign Intelligence Advisory Board. Richard Helms's acceptance of the too-low estimates of enemy O/B produced by MACV and endorsed by DIA was doubtless influenced by his desire not to damage his equities with the military on other issues; a similar consideration was his need to keep DIA aboard the joint CIA-DIA assessments that told the Johnson Administration its bombing campaign was not seriously damaging the enemy's capabilities or will to persist. The O/B positions insisted upon by MACV and the Saigon Embassy were dictated by the perceived need to avoid the public relations damage a suddenly larger O/B estimate would cause. And Walt Rostow's belief in and dedication to the prospect of victory led him to gloss over such troubling intelligence as Saigon Station's pre-Tet warnings, which ran against the administration's public affairs campaign to bolster flagging confidence in its Vietnam strategy.

A principal reason why CIA's data and judgments may have had so little influence on policymaking was that decisions on what to do in Vietnam were not being made in a political vacuum, but had to be developed by leaders whose political party had long been accused of "losing" China and not winning in Korea. Presidents Kennedy and Johnson each affirmed repeatedly that he was not going to be the President who lost Vietnam and the rest of Southeast Asia. Told by CIA that "you can't win in Vietnam," they might well have told themselves, "I can't not win in Vietnam."

Finally, these episodes illustrate yet again the dilemma DCIs and senior intelligence professionals face in cases when they know that unvarnished intelligence judgments will not be welcomed by the President, his policy managers, and his political advisers. At such times intelligence officers must decide just what balance to make between the ever-present contradictory forces of whether to tell it like it is (and so risk losing their place at the President's advisory table), or to go with the flow of existing policy by accenting the positive (thus preserving their access and potential influence). In these episodes from the Vietnam era, we have seen that senior CIA officers more often than not tended toward the latter approach. DCI McCone did so in early 1963 when he chose the policy managers' interpretations of intelligence over the judgments of his own professional staff. So did DCI Helms in 1967 when he struggled to avoid a sharp dispute with MACV on the O/B dispute. And so did CIA's senior Vietnam professional, George Carver, on many occasions.(7) But on at least one occasion, over a period of months in 1964-65, DCI McCone preferred sticking with his own judgments on how the United States should prosecute the war in Vietnam, refusing to ratify the views of other advisers on the course President Johnson was setting. The result was that his persistent candor left him frozen out of the President's inner circle.

In the end, the story of intelligence and the Vietnam conflict is one of competing forces: many potent influences on policy, some of the most significant of them extraneous to Vietnam, versus the obligation CIA's officers had to present their findings candidly, to try to "tell it like it is." As a working-level participant in some of this history, the author can attest to the frustration CIA officers experience when they find no one listening to them downtown. It may be ever so, when intelligence comes up against committed policymakers grappling with intractable, highly charged crisis situations. But the obligation to present candid intelligence findings still applies. One balm for such frustrations when they do occur is the fact that since CIA's founding, every US President, Republican and Democrat alike, has asked for, received, and often benefited from the input of dispassionate, professional intelligence. It is a safe bet that Presidents will continue to need and even welcome such inputs, whatever their ultimate influence on policy decisions.

It is also the author's view that in these episodes Agency officers performed their greatest service when they maintained CIA's professional intelligence integrity without regard to whether candor would or would not prove congenial to their DCIs and to policymaking consumers. To the degree that CIA officers withheld or modified their judgments, they were not only distorting intelligence but also undercutting CIA's very raison d'etre.

There remains an underlying question of whether all the events in this study of producer-consumer relations were taking place within a context of foreordained US failure in Vietnam. Nothing is inexorable; given much stronger South Vietnamese administrations, an earlier and more determined "Vietnamization" effort, and a sharper sensitivity among US policymakers that the war's outcome hung more on political considerations than on body count, the outcome in Vietnam might have been different. Nevertheless, the basic necessity for victory was probably a total American determination as fierce as that of the enemy's to sacrifice and persevere. Successive US administrations and Congresses, and American society at large, were unable to sustain such a degree of determination.

Finally, US policymakers could have acted more wisely, and might have had more successes in Vietnam than they did, had they been more receptive to more of the bald facts and probing interpretations CIA analysts gave them along the way. But the war's outcome was determined by historical considerations far broader than those examined here, and no firm estimate can justifiably be formed of what weight and impact CIA judgments had--or could have had--among the sum of those considerations.

Footnotes

(1) The author recognizes that gauging such intangibles as the quality of intelligence and its impact is an inexact science, and is also the refracted product of the particular gauger's lenses.

(2) That CIA's doubts ultimately contributed to Secretary McNamara's fading vision of light at the end of the tunnel is illustrated in his In Retrospect, where he publishes (pp. 321-323) a list of his and others' misconceptions about the Vietnam War; certain of these match counterarguments CIA officers had long tried to sell higher authorities.

(3) Equally to his credit, Carver not only continued to have considerable impact with top policymakers in the years following those studied here, but also is credited with having generally contributed candid intelligence inputs during that period.

(4) It should be noted, however, that despite its rigorous methods, ORR, joined by other CIA analysts, later discounted and underestimated the magnitude and significance of North Vietnamese support reaching the Viet Cong through Cambodia's port of Sihanoukville.

(5) Mr. McNamara italicizes this concluding phrase from the assessment. In Retrospect, pp. 292-294. He adds that he did not see this unique Agency document until discovering it in Johnson Library files in the course of researching his book.

(6) In Retrospect, p.294.

(7) As we have seen, however, it is notable that as an O/NE analyst prior to becoming CIA's chief Vietnam affairs officer, Carver had been a champion of candor, whatever the views of policymaking consumers; and that at the end of March 1968 he did give the White House the kind of bad news that earlier he had often played down.