MEMORANDUM FOR THE DIRECTOR FROM ED PROCTOR

• Document Type: FOIA
• Collection: FOIA Collection
• Document Number (FOIA) /ESDN (CREST): 03386239
• Release Decision: RIPPUB
• Original Classification: U
• Document Page Count: 11
• Document Creation Date: July 13, 2023
• Document Release Date: December 15, 2022
• Case Number: F-2022-01326
• Publication Date: January 18, 1974

Approved for Release: 2022/10/17 C03386239
,
MEMORANDUM FOR: The Director
Execud7o
Attached for your background information
is a memo prepared in OSR that puts in per-
spective the recent columns by Joe Alsop on
Soviet Defense spending.
The memo refers to "tabs" occasionally.
I have not included these in order to spare
you unnecessary detail.
Ed Proctor
18 Jan 74
(DATE)
0
FORM NO. 101 REPLACES FORM 10.101
1 AUG 54 WHICH MAY BE USED.
(47)
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
lairmer
MEMORANDUM FOR: Director of Central Intelligence
FROM : Deputy Director for Intelligence
SUBJECT : Manpower Costs as a Share of Total
Defense Expenditures
1. The data I provided as backup for your
briefing and the numbers in the memo I sent you
concerning the Alsop articles are both consistent
and correct. The problem--you will remember my
exchange with Senator Byrd on this issue--revolves
around the definition of manpower costs selected.
2. In paragraph 8 of the Alsop memo, the
33 percent share for US manpower costs includes
only the pay and allowances for active military
personnel. When all personnel costs are included--
expenditures on retired personnel, pay for civilian
employees, family allowances and the like--the
percentage can range from 50 to 60 percent de-
pending on the source consulted and the definitions
selected. For our present estimates, OSR is using
the Five Year Defense Program (FYDP) dated
January 1973. These data will shortly be up-
dated when the January 1974 FYDP becomes
available.
3. By the way, you are correct--the Soviet
manpower costs cited in paragraph 8 of the Alsop
memo are for the 3.8 million active military per-
sonnel estimated to be in the Soviet armed forces
in 1973.
(b)(3)
(b)(6)
EDWARD W. PROCTOR
Deputy Director for Intelligence
(b)(3)
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
_
Approved for Release: 2022/10/17 C03386239
Nive
January 1974
MEMORANDUM
Comments on Criticisms by Joseph Alsop
of US Intelligence Estimates of
Soviet Defense Spending
1. In several of his recent columns appearing
in the Washington Post--specifically on 14 November
1973 and 9 and 11 January 1974--Joseph Alsop charac-
terizes US intelligence estimates of Soviet defense
spending as being grossly understated. He also makes
much of the fact that the USSR has a much lower
military pay bill than the US does and so is able
to spend a greater portion of its defense budget
for the procurement of military hardware.
2. Mr. Alsop's basic message actually has little
to do with economics: It is the statement that "we
have been fooling ourselves blind, and for years on
end, about the scope, intensity, and general success
of the Soviet defense effort." This thesis cannot be
proved or disproved through analysis of economic data.
US estimates of Soviet defense programs and activi-
ties flow directly from information on the forces
themselves--they are not derived from economic data
published by the Soviets or estimated by US intelli-
gence officers. The various monetary measures CIA
uses to illustrate the economic implications of the
estimated forces are themselves derived from the
observed physical activity. Mr. Alsop compounds the
basic fallacy by a vitriolic--and inaccurate--attack
on the economic data, an attack which largely dis-
credits itself in terms of logic and elementary
economics.
Comments and queries regarding this memorandum are
welcomed. They may be directed to
of the Office of Strategic Research,
(b)(3)
(b)(6)
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
- p.4111�
3. To substantiate his argument, Mr. Alsop
offers the following major points*
Lower Cost - of ' Soviet' Military Manpower
Soviet manpower costs are on the order of
$18 billion out of total defense outlays of
$80 billion, leaving $62 billion "to buy real
strength in the form of weapons." The US, on
the other hand, has only $35 billion available
from a $73 billion defense budget to buy weapons.
Estimated Cost of Soviet Border Build-Up
Revised Upward
Revised US intelligence estimates of the
cost of the Soviet build-up on the China border--
recently "tripled" over previous estimates--
are "still inadequate because some omissions
were not corrected."
Cost Estimates of Major' Soviet Programs
� Patently Low
If one examines official US estimates of the
dollar cost of selected Soviet defense programs
about which the US knows a good deal--specifically
the deployed SA-5s and SA-3s, the command communi-
cation network, and the test and development
programs for the latest generation of ICBMs--
"our estimates of Soviet defense spending simply
cease to add up in US terms".
Soviets Acknowledge Defense Gets Large Share
of GNP
Estimates by two Soviet economists--corroborated
by the prominent Soviet dissident scientist,
Andrei Sakharov--place Soviet defense spending
at 40-50 percent of GNP compared to US intelli-
gence estimates of only 9 percent.
The complete texts of the three columns are attached
at Tab A.
-2-
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
Imeri NO,
4. Mr. Alsop's contentions and conclusions
reflect a variety of misunderstandings and misuses of
US intelligence estimates of Soviet defense spending.
One confusion that appears to underlie a number of
Mr. Alsop's judgments concerning analyses by US
intelligence is that estimates of Soviet defense out-
lays in ruble terms can be used interchangeably with
estimates of the dollar cost of Soviet programs. In
fact, the two measures are quite different in concept
and have distinctive applications.
--The estimated dollar costs of Soviet
programs represent what it would cost the US
to purchase identical military equipment, hire
the same number of people with like skills and
carry out the RDT&E, and operations & maintenance
programs in the same fashion as the Soviets.
Conceptually, the Soviet forces are viewed--
for costing purposes--as alternative US forces.
--Estimates of the ruble costs of Soviet
programs are based on our knowledge of the
Soviet economy. The ruble figures attempt to
reconstruct the defense effort of the USSR in
Soviet financial terms, i.e., as they would
appear to Soviet defense and economic planners.
5. The estimated dollar costs of Soviet defense
programs are frequently compared with US defense
figures. Such comparisons are not appropriate for
drawing inferences concerning the relative produc-
tivities of the Soviet and US economies or for
comparing the internal distribution of resources
in the two countries. Moreover, comparisons of
US defense spending and the estimated dollar costs
of Soviet defense programs cannot be used to draw
valid conclusions about the relative military ef-
fectiveness or capability of US and Soviet forces.
6. Estimates of Soviet defense spending in
rubles form the basis for analyzing the magnitude,
direction, and resource implications of defense
programs from the perspective of the Soviet decision-
maker. Such analyses give some appreciation of the
relative priorities that Soviet decisionmakers
probably attach to individual defense programs as
well as to the defense effort as a whole.
-3-
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
Nor fter - - NeOlY
'Lower: Cost: of Soviet' Military Manpower
7. Mr. Alsop apparently mixed ruble and dollar
estimates in deriving the estimate of $18 billion
for Soviet military manpower cited in his 11 January
article. His starting point was a reported statement
by Brezhnev that military manpower costs are currently
22 percent of total defense outlays--CIA estimates the
share to be about 25 percent in 1973. Both percentages
are based on ruble data, reflecting the low cost of
Soviet manpower in' rubles. Neither figure represents
what Soviet manpower would cost in dollars as a share
of the total Soviet defense effort valued in dollars.
8. An estimate of the share of the Soviet de-
fense effort devoted to personnel based on dollar costs
of all Soviet programs can be derived by using the
dollar costing methodology described above. This
figure is only appropriate, however, for viewing
the situation from the standpoint of a US defense
planner considering the present Soviet force as an
alternative US force. When active Soviet manpower 3,
is costed at US pay rates, it amounts to about $35
billion in 1973, of a total Soviet defense effort
of about $81 billion in that year. Hence, the man-
power cost share of this alternative US force--in
dollar terms--is about 43 percent, not 22 percent or
25 percent. Comparable US manpower costs in 1973
were on the order of $24 b i n out of a total
some $73 billion, or about 33-6Tce"n7n--
9. Mr. Alsop also makes the common false
assumption that all funds not expended on personnel
are available for buying new equipment. RDT&E and
operating & maintenance costs must be met as well.
To. determine the relative sizes of the US and Soviet
military procurement efforts, the procedure used
by CIA is to estimate the dollar value of individual
Soviet military production programs as if they were
carried out in the US, sum them, and compare the
result with total US procurement outlays. The
results of such a comparison show that the esti-
mated dollar values of the military procurement of
the US and the USSR in 1973 are virtually identical--
approximately $18 billion. The reason for this
-4-
7 I
Approved for Release: 2022/10/17 003386239
Approved for Release: 2022/10/17 C03386239
low limo 1) . , %or lie
unexpected result is quite simple, if counter-
intuitive: although manpower is relatively more
expensive in the US than in the USSR, hardware is
relatively cheaper.
Estimated Cost of Soviet Border Build-Up Revised
'Upward
10. Mr. Alsop's 14 November column cites alleged
changes in official US estimates of the cost to the
Soviets of the Sino-Soviet border build-up as evidence
of the failure to assess the situation correctly--at
least at the time of the earlier estimate. The later
figure is significantly higher because it is an
estimate of a different set of activities, over a
longer time period, and using a different price base:
--The earlier estimate reflects cumulative
incremental expenditures to increas the border
force and does not include cos or strategic
attack, strategic defense, and border guards.
The later estimate includes costs to increase
the force, costs to maintain those forces present
before the build-up began, costs for strategic
attack and defense and border guards, and costs
reflecting improvements in order-of-battle
and facilities estimates.
--There is a difference in the span of years
covered by the two estimates (the earlier esti-
mate covers 5 years, the later one covers 8).
--The price base was moved forward between
the two estimates (the earlier estimate is in
1968 dollars while the later is in 1972 dollars).
Further, the allegation that "a series of critical,
vastly expensive factors had been omitted from the
old $6 billion price tag" (including "30 brand new
-Soviet jet airfields") is incorrect. For the same
set of forces over the same period of time, the
present estimate is in close agreement with the
previous estimate.*
4A more detailed comment on the Alsop column of
14 November is at Tab B.
-5-
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
lior few Nor vor
, .
Cost. Estimates ' of ' Major- Soviet Programs
'Patently -Low
11. Mr. Alsop implies that estimates of the
dollar costs for such defense programs as air defense
systems, development of strategic missiles, and
command communications are obviously too low. This
criticism appears to be based in part on a misap-
prehension of what the estimates of the dollar costs
represent. After these programs are defined and
measured using all available intelligence information,
estimates of what it would cost the US to reproduce
these exact programs are made--in dollar terms as
described above. These costs are, of course, in
error to the extent our assessment of Soviet weapons
systems is incomplete or incorrect. This type of
error, however, is not likely to produce grossly
misleading estimates.
� Soviets Acknowledge Defense Gets Large Share
of GNP
12. Mr. Alsop cites a 1971 -samizdat article
by two Leningrad economists as evidence for his
contention that the USSR has a burden of defense
far greater than estimated by US analysts. CIA
analysts did a critical review of this monograph
when it became available in the West early last
year.* This examination uncovered faulty assump-
tions and techniques in the Golitsov and Ozerov
article which made the conclusions of their study
completely invalid. The samizdat paper adds
nothing to the understanding of the size of Soviet
national income or Soviet defense spending.
--No new basic data on these subjects
are presented in their paper. It uses Soviet
published statistics long available to the West.
A translation of the sami'.2dat article and a
fuller critique of it are at Tab C.
7
,
Approved for Release: 2022/10/17 C03386239
Approved for Release: 2022/10/17 C03386239
Now" Noe,'
"
--The key assumptions, statistical methods,
arbitrary adjustments of the basic data and
simplistic approach are so faulty as to dis-
credit the results.
Sakharov repeats rather than confirms the assertions
in the samizdat paper.
Attachments:
Tab A
Tab B
Tab C
-7--
Approved for Release: 2022/10/17 C03386239
No/
TO:
Approved for Release: 2022/10/17 C03386239
Routing Slip Now
ACTION
INFO.
ACTION
INFO.
1
DCI
_11
IG
2
DDCI
12
D/PPB
3
DDS&T
13
SAVA
4
DDI
14
ASST/ DCI
5
DDO
15
AO/ DCI
6
DDM&S
16
EX/ SEC
7
D/DCI /IC
17
8
Di ONE
18
9
GC
19
10
LC
20
SUSPENSE
Remarks:
Remarks:
Date
DC1/11=
Approved for Release: 2022/10/17 C0338623s9 /