325. Telegram From the Special Assistant for Vietnamese Affairs (Carver) to Director of Central Intelligence Helms 1

CAS 1983. For Knight only from Funaro.2 Ref: A. Saigon 1925 [less than 1 line of source text not declassified], B. Saigon 1926 [less than 1 line of source text not declassified].3

Circle now squared, chiefly as result of Westmoreland session (and perhaps Komer dinner). We now have agreed set of figures Westmoreland endorses. Mission seems on verge of successful conclusion, though final t’s to be crossed tomorrow.
Komer dinner was relaxed, frank, cordial and (most usefully) private. We reviewed whole estimate exercise. I gave blunt outline of weaknesses in MACV case, of peculiar MACV behavior, and of fact that agreement impossible unless climate improved enough to permit serious discussion. Komer did not agree with our substantive position and repeated some caustic remarks about the estimate but he did listen.
On morning 13 September, General Davidson advised me Westmoreland meeting would do nothing but formalize our impasse, since Westmoreland would never accept our position. When meeting convened (with Abrams, Komer, General Sidle (PIO)4 and INR reps attending), Davidson gave rather biased account of proceedings, noting our impasse on figures, saying he thought our paragraph written to avoid quantifying irregulars (last para Ref A) unacceptable and outlining his draft cable by which General Westmoreland could advise General Wheeler of our inability to agree. Komer weighed in with replay of his thesis, recommending acceptance MACV position but acknowledging logic in some of Washington views. I then reviewed history and context whole estimate, the Saigon discussions, and the rationale behind each of the joint Washington representative figures, and (as tactfully as possible) the way the procedure through which national estimates produced reduced our negotiating latitude. I also took up “quantification paragraph,” indicating that Davidson had quoted out of context and showing why we thought it met both Washington [Page 800] and MACV needs. I concluded with general remarks tracing history of U.S. military estimates on Vietnam since 1956, noting that consistent record of underestimation plus recent (i.e. McChristian)5 methodology which required constant retroactive juggling had contributed mightily to our credibility problems, that we needed baseline which would not have to be adjusted again, and that when in doubt we well advised, from all angles, to err on side of caution rather than optimism.
Westmoreland most cordial and receptive. Said he agreed with most of my observations and could see the clear logic behind both sets of figures, which were really not that far apart. He also saw the rationale behind our “irregular quantification” paragraph and had no problems with it, though he would want to take a final look at it in writing. He asked if I would convene both the Washington and MACV analysts to review the evidence once again and see if we could resolve our differences.
When the analyst meeting was convened (with Davidson represented by his deputy) I took Westmoreland at his word, usurped the chair, and announced that all constraints on totals were off and we could settle down to serious discussion of evidence and issues. During about four hours of brisk discussion we hammered out the following set of agreed figures, which General Westmoreland has already endorsed: main and local force 119,000, admin services 35,000–40,000, guerrillas 70,000–90,000 for military total spread of 224,000–249,000. Political 75,000 to 85,000.
On the whole, I think we can live quite comfortably with the above figures (which the DIA team and INR rep endorse). We have no dispute on the main and local force 119,000 figure. The text of the operative sentence in our admin service paragraph (see Ref A) now reads “In light of these considerations, we estimate that there are now at least 35,000–40,000 administrative service personnel who are performing essential administrative support functions full time.” The rest of the paragraph is unchanged. We gave a little cosmetically but I think this preserves the essence of the judgment in the 14.3 draft.6
There has been some adjustment on guerrillas, but the new figure (70,000–90,000) appreciably lifts MACV’s previous total (65,000) and has the same median (80,000) as the spread in the 14.3 draft.
We gave a little on the political figure, partly to keep MACV on the reservation with respect to the guerrillas, partly because MACV did have a case on double counting (some bodies in both military and political figures, though this was not the reason MACV whacked the [Page 801] figure yesterday), and primarily because the discussion and evidence convinced me that this is not a very good figure anyway. Our present definitions are not adequate or sufficiently precise, we include much more than the real “leadership” and exclude many (e.g. security elements) of whom formal cognizance should be taken.
On above, with endorsement Messrs. Hyland and Moor7 and DIA reps and concurrence General Davidson, I am initiating major study to refine our political categories and hence improve our political holdings. To this end, Mr. Adams will remain Saigon to go over whole subject in detail with MACV analysts and our ICEX officers. Further work will then be carried on in Washington under Mr. Moor’s aegis.
Our agreed figures and irregular quantification paragraph being given General Westmoreland in writing tomorrow. Once he adds signature to already expressed verbal approval our mission successfully completed. I have 1500 hours appointment with Ambassador Bunker on 14 September and shall give him full report.8
General Westmoreland has requested I work with General Davidson and General Sidle to prepare scenario for press backgrounder. I have accepted your concurrence. Please advise.9
On returns, early reservations out of Saigon almost impossible obtain. Subject your approval (please advise immediate) I have authorized Messrs. Moor and Hyland use their present 14 September PanAm reservations. I have reservation for 16 September but will not leave until you concur. Request our wives be advised of these travel arrangements.10
  1. Source: Central Intelligence Agency, DCI (Helms) Files, Job 80–R1720R, GAC Files. Secret; Immediate Director. The telegram was an attachment to a memorandum from Carver to Helms, November 28, 1975.
  2. Knight was Helms’ pseudonym; Funaro was Carver’s.
  3. See Document 314 and footnote 8 thereto. A typed note added at a later time notes that these two cables were “messages to Knight concerning the VC strength estimates.”
  4. Brigadier General Winant Sidle, Chief of MACV’s Public Information Office.
  5. Major General Joseph A. McChristian was the former MACV Chief of Intelligence.
  6. See footnote 8, Document 314.
  7. Dean Moor, an expert on North Vietnam in the CIA’s Office of Current Intelligence.
  8. In his meeting with Bunker, Carver reported on the agreement with MACV on the figures but left out “details now better forgotten” pertaining to the negotiations that led to the coordinated estimate. (Telegram CAS 2043 from Saigon, September 14; Central Intelligence Agency, DCI (Helms) Files, Job 80–R1720R, GAC Files)
  9. In telegram CAS 1988 from Saigon, September 13, Helms responded: “Have no objection to your undertaking what General Westmoreland asks, but do not feel that such press backgrounds should be related in any way to NIE process. Will make determination only after you return here whether or not estimate will be issued.” (Ibid.)
  10. The delegation left Saigon as scheduled.