WILLIAM T. LEE'S CRITICISM OF CIA - DIA ESTIMATES OF SOVIET MILITARY SPENDING

• Document Type: CREST
• Collection: General CIA Records
• Document Number (FOIA) /ESDN (CREST): CIA-RDP87T01145R000200280017-8
• Release Decision: RIPPUB
• Original Classification: C
• Document Page Count: 12
• Document Creation Date: December 27, 2016
• Document Release Date: May 2, 2012
• Sequence Number: 17
• Publication Date: May 19, 1986
• Content Type: MEMO

Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
Memorandum for:
FROM:
D/SOYA
Attached is Jim's memo on the
article. Frankly, Soviet defense spending grew at an average
percent
period--including a five-year estretchywhen GNP
grew at about 2 percent per year--has done more
damage to his credibility than anything we could
say.
STAT
STAT
STAT
reader regarding the extentItoewhichnhis7figures
are "according to both Soviet economic plans and
subsequently published data..",
tell , the reader that it is the He doesn t
0
interpreting these data that method of
astounding growth rates. produces his
Douglas1J. MacEachin
Director,
SOYA
Office of Soviet Analysis
STAT
STAT
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
i X1
19 May 1986
MEMORANDUM FOR: Deputy Director for Intelligence
THRUUGH:
FROM:
SUBJECT:
REFERENCES:
Douglas J . MacEachi n
Director of Soviet Analysis
Chief, Defense and Economic Issues Group/SOVA
Criticism of CIA -
DIA Estimates of Soviet Military Spending
A. John McCaslin, "U.S. Intelligence on Soviets Faulted,"
The Washington Times, May 19, 1986, pp. 1A, 1OA.
B. I "Soviet Military Spending Still Growing,"
National Security Record, No. 91 (May 1986), p. 5.
1. This morning's issue of The Washington Times includes an article
criticizing the estimates of Soviet military spending'
pen ing recently presented by
CIA and DIA in joint testimony before the Joint Economic Committee of the US
Congress (Reference A, attached). The Times article, in turn, is based on an
article of DIA in the May issue of the Heritage Foundation's 25X1
National Security Record (Reference B, attached). 25X1
2. Although recent article provides few specifics on the basis 25X1
of his disagreement with the CIA-DIA estimates, his position appears to be
based on much the same arguments that he has advanced in other publications
over the past two decades. In particular, his assertion that CIA and DIA are
underestimating Soviet defense spending is evidently based largely on a
comparison of the results of the CIA-DIA "direct costing" estimates of Soviet
spending for military procurement with the higher estimates that he obtains
through analysis of unclassified Soviet statistics on the output of the
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
2bX1
machine-building industries (which produce military hardware, consumer
durables, and investment goods) and its distribution to non-military
3. We have recently completed a paper assessing) approach and
concluded that it is extremely misleadin Our paper notes that, depending on
how the Soviet datd are interpreted method may yield results much
closer to the CIA-DIA estimates than to those that he obtains, but that it is
too fraught with uncertainty to be used in estimating the value of Soviet
spending for procurement.
25X1
25X1
25X1
25X1
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
25X1
SUBJECT: Criticism of CIA -
D IA Estimates of Soviet Military Spending
UI/SUVA/DEIG/DEA/DEP
Distribution:
Orig - Addressee
1 - D/SUVA
1 - DD/SUVA
2 - C/DEIG
1 - C/UEIG/UED
1 - C/DED/EI
1 - C/DED/DEP
(19 May 1986) 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
C WrnMjt!qTUU ~ .Ct
U.S. inteffigence
on Soviets faulted
By Joan MCCas in
THE WASHINGTON TIMES
A Defense Intelligence Agency
anal~'st says both the DIA and the
Central Intelligence Agency have
seriously underestimated Soviet de-
fense spending.
In an article in the May issue of
the The Heritage Foundation's Na-
tional Security Record. DIA analyst
11'.T Lee says that. contrary to DIA
and CIA estimates of relatively flat
Soviet military spending, the Soviet
Union has been spending a growing
percentage of its total state budget
for military purposes.
A CIA and DIA report on Soviet
SOVIET
From page ]A
spending, despite General Secre-
tary Gorbachev's public promises of
economic improvement," Mr. Lee
Wrote.
Citing both Soviet economic plans
and subsequent published data, Mr.
Lees report says Soviet military out-
lays grew from about S6- billion in
1970 to about 5207 billion in 1983,
and probably exceeded S229 billion
in 1985. (The dollar figures are
based on exchange rates which tend
defense spending presented to Con-
gress in March said Soviet spending
on weapons has been constant since
1975 "and is likely to grow little. if
any, over the next five years.'
But Mr. Lee's article takes issue
with those estimates.
It says Moscow's military expen-
ditures are increasing at about 8.5
percent annually. This compares to
a rate of less than 3 percent for U.S.
defense expenditures over the same
period.
"There has been no indication to
date of any significant reallocation
of resources to non-military
see SOVIET, page IOA
to understate the size of the Soviet
spending.)
"And these estimates of total So-
viet military outlays are minimal be-
cause they do not include the cost of
civil defense. military pensions. and
probably other military-related ex-
penditures;' he says.
Mr. Lee adds that when weapons
procurement is separated from
other militan, expenditures, the an-
nual growth rate of Soviet weapons
procurement alone averated more
than 9 percent per year for the past
10 years.
the official CIA estimate of less than
one percent per year," he says.
From the late 1960s through 1981,
Soviet rnilitan. outlays accounted
for about one-third of the state bud-
get each year. Since 1982, Mr. Lee
says. the military's share of the So-
viet budget has risen. and in 1984
and 1985 was about 38 percent of the
total.
"This constant shift in Soviet na-
tional priorities to the military has
not been accidental;' he says. "Soviet
leaders planned it that way. as can be
seen by the priorities of their three
five-year plans covering the period
from 1971 through 1985"
Since the mid-1970s, the Depart-
ment of Defense has reported a
trend toward increased complexity
and advanced technology in new So-
vie: weapons systems, the full cost
of which "is reflected in constant So-
viet prices" says Mr. Lee.
He says it is likely that at least
three-quarters of the growth of So-
viet military expense in the last dec-
ade is the result of accelerated tech-
nological advances, including
increased missile mobility and accu-
racy, introduction of AWACS radar-
command aircraft, and new fighter
jets that provide "look-down, shoot-
down" capabilities.
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
A Report on the Congress and National Security Affairs
THE HERITAGE FOUNDATION ? 214 MASSACHUSETTS AVENUE. N.E. ? WASHINGTON, D.C. ? 20002 ? [phone]
NO 91
In this issue ...
Reagan is Reagan, p. I
\tter toe scars in 01iuc. Ronald Re,igan is aciivels pursuing the
n;ui"ni,il seuunts agenda he stood for "hen he ran for the
hr(',idcnu\
I)efendirtg \1A1). p. A
hs Daniel 0. (.raham
Vs hs d? loaner defense secretaries McNamara and Brown persist
if! siclending \I -AD and opposing SDI.'
Susiet Alilitar Spending Still Growing, p. 5
bs \\.1. 1Xc
IcteilIV euce an.ilsst lee sass th;it C I:A estimates of Soviet military
spcnoin,i ire to,i loss
Insiders Report, p. 4
-RE\A\10 Opens Vsashington Oflice
- Needed: a Sovietologist for the \S(
Baldrige Boosting Socict Trade
Is the C \ Going Broke'.)
Downgrading Public Diplomacy
-Cranston on kkar Passers
--(iorbachec Building Bridgeheads
--The Navy Cuts Costs
MAY 1986
Getting Loose from the Salt II 'far-Bahl, p. 6
Can Ronald Reagan ever get out of the aereenient he
campaigned against in 1950'The betting i, that he swill,
before the end of this sear.
Reagan is Reagan
The Reagan Doctrine may have been conceived earlier, but
it became reality in 1986. The air strike that Reagan ordered
against Libya's Qaddafi last month was only the most recent
and most dramatic in a series of moves that suggest that the
President finally has taken firm control of national security
policy On most issues, Reagan now is actively pursuing the
agenda he stood for when he ran for the presidency in 1980.
Consider a fevv of the recent national security actions or
decisions by the Reagan administration:
? A request to Congress for 5100 million for the Nicara-
guan freedom fighters, including S70 million in military
assistance, and the use of the full weight of the office of
the president to gain Congressional approval.
? A vvillingness by the President to greet Angolan freedom
fighter Jonas Savimbi in the White House and to promise
him L .S. military assistance in his struggle against So-
vier /Cuban control of Angola.
? An increase in meaningful assistance to the Afghan free-
dom fighters.
? A decision to send U.S. Stinger anti-aircraft weapons to
the freedom fighters in Afghanistan and Angola.
? ,A review of ways to provide assistance to the resistance
groups opposing the Soviet-style regime in Ethiopia.
? The use of L.S. Navy planes to capture the Achille Lauro
hijackers over the Mediterranean.
? The exercise of freedom of the seas in the Gulf of Sidra
by the L.S. Sixth Fleet, including the sinking of Libyan
naval craft and air attacks on radar sites.
? The issuance of an order sharply reducing the number of
Soviet officials in Ncs+ York.
? A decision to cut the number of Soviet pc:-sonnel ssorking
at the U.S. Embassy in Moscow.
? The restriction of other communist officials from travel-
ing outside the New York City area without advance
notification.
? No significant unilateral U.S. concessions at the Geneva
arms control negotiations.
? No concessions to Gorbachev's demands for a morato-
rium on nuclear testing.
? No concessions to Gorbachev's persistent efforts to block
or limit the Strategic Defense Initiative.
? No concessions to the efforts by Moscow (and some
Members of Congress) to stop deployment of a U.S. anti-
satellite weapon.
? No concessions to get a second summit meeting.
? No significant lessening of the military modernization
program.
? The continued issuance of official government reports of
Soviet arms control violations.
? Movement toward proportionate responses to Soviet
arms control violations.
? Movement toward an end to the SALT 11 agreement.
? Pressure on Marcos in the Philippines and Duvalier in
Haiti to step down peacefully, avoiding violence and pos-
sible takeovers by leftwing forces.
A Positive New Policy
All this adds up to a new, positive policy of containing,
confronting and ultimately reversing the tide of Soviet imperi-
J
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
Soviet Military Spending Still Growing
by W. T. Lee
In March rile Central Intelligence Agency and the Defense
Intelli,Lcence . Igenc?l jointly presented a new report on Soviet
dcfen le cpcnuling to the Joint Lconunnic Committee of (,on-
Ill(- report claims that Soviet weapons procurement
has been constant sine 0- 5 and is likelh to gros little, it
ant. eier the nest five rears. This C/.-i/DIA estimate has
linen anuruuttlion to the opponents of t S. defence spending.
In l07 the (1.'I revised upward its estimates of Soviet
nrilitart spending, nearly doubling its previous estimates.
lilt, lollop, ing rear, in 1 0,77, longtime defense intelligence
analyst If illianr I. Lee krone an article for the National
.Strategy Information Center entitled Understanding the So-
ytet A1ilitare Threat. In it, A1r. Lee contended that the revised
( L-1 estimates still understated the cost of Soviet weapons
proerams and that since the CIA had not changed its nreth-
odo/o r 6ir estimating .Soviet expenditures, its underesti-
niates o%ould continue..\os, nine rears later. Mr. Lee finds
theft his prediction gas correct. The CIA, he claims, ''has
performed rather precisely as predicted- in continuing to
urrdrrestinntr .Sorirt military spen(ling. Mr. Lee explains his
mu, h higher estimates in the- follolsing article:
During the past three years a number of observers have
argued that Soviet military spending has grown very slowly
since 11)'0_ Basing their conclusions on CIA analyses, they
claim that Soviet military expenditures increased only about
tcvo percent per year over the past ten years, while weapons
procurement did not gross at all - at least through 1983.
Iloccever, estimates of Soviet military spending that are
b,r>ed on Soyict economic statistics tell an entirely different
stun- According to both Soviet economic plans and subse-
- ------ ----- ----
cluent_Published data, Soviet military outlays grew from about
50 billion rubles in 1970 to about 154 billion in 1983, and
probable exceeded 170 billion in 1985. This amounts to an
annual :tccragc growth rate of about 8.5 percent, compared to
a rate of Irss than 3 percent for U.S. defense expenditures
Oyer the santc period. And these estimates of total Soviet
military outla s are minimal because they do not include the
cost of civil defense, military pensions, and probably other
ntilitare-related expenditures as well.
\\ hen cccapons procurement is separated from other mill-
tarc expenditures. the annual growth rate of Soviet weapons
procurement alone averaged more than nine percent per year
for the past ten years. This is considerably higher than the
olllci.tl CIA estimate of less than one percent per year.
The nnilitarr burden on the Soviet economy, reckoned as a
share of GNP, rose from 12 percent in 1970 to 18 percent in
11),M) and probably reached 21 percent in 1985. as the Soviets
reckon constant 1970 prices. Even measured in 1985 prices.
the Soyict military burden last year was about 17 percent of
G\I'. or more then tcyo and a half times the comparable U.S.
figure for 1955 (about 6.5 percent of GNP).
I rem the late I )6Os through 1981- Soviet military outlays
accounted for about one-third of the state budget each year.
Sink 1982, the military's share of the Soviet budget has risen,
.and in the rears 1984-85 ryas about 38 percent of the total.
I his constant shift in Soviet national priorities to the military
its not been accidental. The Soviet leaders planned it that
'S.+\. as can be seen be the priorities of their three five-year
plans coyerint, the period from 1971 through 1985.
All of the independent check-points from Soviet sources
have confirmed my estimates. While the official U.S. govern-
ment estimates were revised upward in 1976 by factors of tcsa
to four, mine have not changed. These are, therefore, high
confidence estimates of Soviet military outlays and of the
military burden on the Soviet economy.
Since the mid-1970s, the Department of Defense has re-
ported a trend toward increased complexity and advanced
technology in new Soviet weapons systems. The full cost of
this trend is reflected in constant Soviet prices. From 1962 to
1972, when the pace of change in Soviet weaponry was rela-
tively slow. the cost of technological advances was increasing
Soviet procurement costs by 6 to 7 percent per year.
Thus, it is likely that at least three-quarters of the growth of
Soviet military procurement costs in the ten years since 1976
is the result of accelerated technological advances, including
such things as the increased accuracy and mobility of Soviet
missiles, the introduction of .A''NACS, and ness fighters with
advanced avionics that provide look-down shoot-down ca-
pabilities. Further evidence that the cost of technological im-
provements is the principal cause of recent increases in Soviet
military spending can be found in the Soviet production statis-
tics contained in Soviet Military Power, which is published
annually by the Department of Defense.
The large increase in the total Soviet budget over
the last 15 years can only be explained by continu-
ing growth in military spending.
In most cases, the numbers of weapons produced in recent
Nears has remained stable or even declined. This is misleading
for systems such as the Typhoon submarine, the new Soviet
aircraft carriers. and other weapons where the physical size of
the individual units has increased substantially. Nevertheless.
the overall pattern is clear: the growth of Soviet procurement
costs is being driven more by improvements in weapons tech-
nology than by increases in numbers of individual weapons.
The growth in the Soviet state budget since 1970 is fully
compatible with these estimates. From 155 billion rubles in
1970. total budget outlays have grown to nearly 400 billion
rubles in 1985. But outlays for investment and social welfare
have grown much less rapidly than the budget as a whole.
accounting for only about 100 billion of the 245 billion ruble
annual increase since 1970. Subsidies may have increased as
fast as the budget. but they remain a small portion of it.
Consequently, the large increase in the total Soviet budget
over the last 15 years can only be explained by continuing
growth in military expenditures, especially for high-tech
weapons systems. The Soviet Union is continuing to spend a
very large and even growing percentage of both its total state
budget and its GNP for military purposes. There has been no
indication to date of any significant reallocation of resources
to non-military spending, despite Gorbachev's public prom-
ises of economic improvement.
UI T Lee is an analyst with the Defense Intelligence Agency.
This article represents his personal views and not necessarily
those of the Department of Defense or the DIA.
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
25X1
Estimating Soviet Military Hardware Purchases:
The "Residual" Approach) 25X1
SUMMARY
Information available as of 28 April 1986 was used in this paper.
It has long been accepted that the Soviet machine-building and
metalworking (MBMW) sector is the source of almost all military hardware--as
well as machinery for consumption, investment, and capital repair. When the
Soviets report data on MBMW output and on the distribution of this output,
however, they do not provide information on the military's share. The secrecy
surrounding this information has led many Western analysts to attempt to
estimate the share from reported Soviet economic data. 25X1
One appealing estimating technique is known as the machinery purchases
"residual" approach. The basic assumption of this approach is that all
military machinery purchases are included in the MBMW output data, but not in
reported purchases. Using this method, analysts subtract the value of
identifiable nondefense purchases from the total output of the MBMW sector.
The remaining output--the residual--is believed to represent the value of
annual military hardware purchases. 25X1
We have conducted a lengthy investigation of this approach. In this
report, we present the results of our attempt to estimate a machinery
purchases residual for 1966-84. To derive the estimate, we reviewed the
available evidence on MBMW output and the estimating techniques used in
previous attempts to apply the residual approach. At each step, we calculated
the uncertainties resulting from various interpretations of the data. I 25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
Because of the great uncertainties associated with the interpretation of
.t -0
the Soviet data used in the residual procedure, we conclude that the CIA
method and two independent methods that were also examined are unreliable as
independent techniques for estimating the level and trend of Soviet military
hardware expenditures. For example, the estimates for the total value of
machinery produced--the starting point for each of the techniques
examined--range from 168 billion rubles to 194 billion rubles in 1980.
The data used in the remaining steps in the analysis are incomplete,
poorly defined, and incomparable in price base and coverage. To estimate the
various categories of nondefense production using these data, for example,
many assumptions must be made that, cumulatively, lead to considerable
variation in the final estimate. The tremendous range in both the levels and
growth rates of residual estimates does not necessarily mean that the methods
are wrong. But they do illustrate a problem inherent in the approach--that
various assumptions and methods used in developing the estimates can cause
widely differing results.
The degree of uncertainty in an estimate of military machinery purchases
calculated by the residual method becomes readily apparent in an analysis of
our estimates. CI,S m3ro.rti UnUCE CF
MWrr IdtuWtm 1 mERr R.A?i*
b. LL~4n OW-rent rI&L..
~d' 1 i B '4 6~~6 O 9 'B 6
ii
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87TO1145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
In current prices they suggest a wide range in estimates of military
purchases--between 4 billion and 15 billion rubles of machinery in 1966 and
between -4 billion and 55 billion rubles of machinery in 1984. Between those
years, the high estimate grew an average of almost 8 percent a year, while the
low estimate declined. The "nominal" estimate--for most steps this is the
mean--grew approximately 7 percent annually, increasing from 10 billion rubles
in 1966 to about 30 billion rubles in 1984. Military machinery purchases
measured in 1970 comparable prices--the Soviet version of constant prices,
which include considerable inflation--grew slightly faster than those in
current prices; the range of uncertainty was about the same.
Our low estimate of military machinery purchases in current prices
actually fell below zero for several years--an intriguing finding since even
the low estimate includes not only residual machinery purchases (any that are
not specifically accounted for), but also a portion of the reported "civilian"
machinery purchases. Therefore, the basic premise of machinery residual
analysis--that all military machinery purchases are included in the MBMW data
but not in reported purchases--may not be true. In our nominal estimate, a
strict accounting of all civilian purchases of MBMW output exhausts the total,
and virtually no residual remains. This suggests two possibilities:
o Some or all military purchases are included in MBMW gross value of
output (GVO) figures but are not hidden in the data as a residual.
Rather, they are distributed among various categories of "civilian"
purchases.
o Some or all purchases of military hardware are excluded from data on
MBMW GVO as well as from reported purchases of MBMW output.
iii
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
 Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8
We are unable to determine which of the hypotheses is true. Because we cannot
estimate what portion of military hardware purchases we capture in a residual
estimate, the technique has little usefulness as an analytical tool.
Even if we were to obtain better definitions of the content of the Soviet
statistics, other problems with the data greatly reduce the value of the
results. For example, even when residuals can be estimated, their levels and
trends are distorted by hidden inflation in the MBMW sector. Official indices
of comparable prices published by the Soviets understate inflation, leading to
an overstatement of growth of real output. As a result, we are unable to
distinguish between real and inflationary growth in the Soviet MBMW sector
using published statistics.
Very little data have been available on the purchases of machinery--
regardless of whether the military or civilian sectors purchase these
goods--since 1972. To produce figures for recent years, we must estimate
values for many of the key variables. If early benchmark estimates of these
values are inaccurate, then extrapolating and using growth indices and planned
growth rates introduce considerable error into the estimates for later years.
25X1
Declassified in Part - Sanitized Copy Approved for Release 2012/05/02 : CIA-RDP87T01145R000200280017-8