CIA HESTOrRiCAL RFv.EW PROGRAM
TITLE: Bconomic Intelligence In Defense Planning
AUTHOR: Clyde C. Wooten
VOLUME: ISSUE: Winter
STUDIES IN
INTELLIGENCE
A collection ol articles on the historical, operational, doctrmal, and theoretical aspects ol intelligence.
AM statements of fact, opinionnalysis expressed in Studies in Intelligence are those of
the authors They do not necessarily reflect official positions or views of the Central Intelligence Agency or any othci US Government enlily, past or present. Nothing in the contents should be construed as asserting or implying US Government endorsement of an article's factual statements and interpretations
Argues that DoD pressure forOn Soviet military outlays and their economic impact haspurious response.
ECONOMIC 1OTELLICENCE IN DEFENSEooten
Id the hurt few years new requirements for mtelhgence clata on costs of present and future Soviet forces and for analysis of the Soviet economic potential with respect to supporting expensive weapon systems have been expressed;
elieveental UW all esttoiaba of Soviet loroe levels be required to meet reasonable tests of eeooomic feasCbutfy. This means that MEl should Include cost estimates and overall budgetary tx^heaQons of the estimated form. It would be very useful to roe to know how the Soviets are aDocatiag their military
McNaanara to the DO,3
We need estimates ofor several reasons. First. It ts very useful for top level planners working on tbe problem of shaping the US defease program to know where the Soviets sir patting theirest, costs to the Soviets give us some asdicanon of the IflreUhood of oertata changes. Fort iropartint to know svhat the Soviet defease budget tlercentage of gross bs tiros! product. And. it 'a also very useful to have ferae feeling for the marginal costs to them of various
changes In their programs
Abb) Eutboveo,3
f course obvious that economicn important constraint on the development of military capabilities.ne thing, however, to recognize that there arc Urnirations on Sovietcapability to rnaintain modern armed forces and quite another to measure that capability for the purpose of testing the feasibility of particular force levels. Or put another way, can "strains" in the Soviet economy caused by advanced weapon programs be effectively measured and the measurement appliedseful way to solution. defense plaiming problems? The purpose of thiso
it the editors' condensationore ojropreheonve study by the author which is available on request
Maaiinc this questionioadci one implied by tbe new DoD cquiieinents for economic intelligence- Given tbe kind of data avail-ible to tbe economic analyst, what kind of response is it feasible for aim to make?
In order to comply with the DoD requests' it is necessary for ioteJligcnoc to develop estimates ofi
L Current Soviet mihtary expenditure allocations within the"
The cost of Soviet forces by mission, mdiidrhg with respect to advanced systems the current and future expenditure: for both present and future systems (be, present operations and maintenance costs, current mvestment for present and future inventories,D costs for future systems).'
The Soviet CNP growth rate or some other measurement of economic capability to support defense expenditures, projected as far as the estimates to be tested are projected.
Allocation of Expenditures
Tbe Soviet military budget Isne-lineingle figure for all military outlays each year. Its interpretation anda job for economic intelligence, is not simplified by the Soviet practice of biding increments to it elsewhere, much as wc hide the CIA budget
hows how this overtly budgeted amount has compared with actual expenditures as estimated by intelligence and with US. obligational authority for defense spending. There is considerable uncertainty associated with the estimated Soviet expenditures, not only present and future but also past, as we shall see. But accepting these figures, we see they give little warrant for extrapolating into the future on the basis of trend. This Is as we might expect; mihtary budgetsroduct of compromise among contrary influences and subject to seemingly unpredictable fluctuations. We are therefore probably not justified in relying on trend analysisechnique for estimating future mihtary expenditures and tbe economic limitations on them.
A considerable amount of analytical ingenuity bas beenin tracking down the bidden Increments of tbe Soviet mihtary budget Data on industrial production have been analyzed to idea-
' Cried and discussed to greater detail bt W. E. Seidell TutcIIigeace for Deiensetudies VIIIl.
tify defense production, largelyesidual method,which are not defense programs in order to isolatebe attributed to defense. Thereumber oftbe residuals approach, but the most important one is thatfigures give bole insight into the mission breakdown ofexpenditures.reakdown has been made anobjective for economic mteuigence. which thereicreto an extensive effort at synthesizing Soviet mibnuy ctWx, in -
particular cents of Soviet weapon systems.
Soofat Com
Meaningful analysis of expenditures rexrulrcs, first, data on prices and quantities, and second, reladonsfalps between these and other prices and quantities. Neither of these conditions Is fulfilled by data directly available oa Soviet military expenditures.
First let us speak of Soviet prices. While It Is an overrimpliScanon to say that prices in the Soviet Union are what someone ays they are, they bear no regular rational internal relationship which couldauds for extrapolation. Because theontrolled and rationed economy, prices areeflection of buyers' and sellers' independent choiceree market. Ruble costs have no necessary rcUtionship to real costs. The variations between the two haw been iiiduectly and approximately expressed by Intelligence (and elsewhere) in terms of divergent ruble-dollar relationships' It may be added thatome divergence among estimates of these divergent relationships. We shall return to this matter; but for the momentecessary only to noteuble In tbe same senseollar.
This circumstance poses tbe first of two difficulties In the costing of Soviet military forces. Because Soviet costs derived from Soviet prices aad quantities arerue reflection of real costs,uite bard to make simple comparisons between the costs of different elements of the Soviet forces (either investment costs or, even rnore difficult, total cost of operations, maintenance, research andtest and evaluation,f course theultiplied when one attempts to. and Soviet rnibtary costs. But precisely this kind of comparison has to be made in order to estimate
' Sea, fee eiirople. the <fccussioa in Alan B.Cottiur. Nuclear Propiau" onf this issue, especially footnote 7.
tbe cost ot Soviet elements in tbe first place. Here arises the second and more serious costing difficulty.
Inasmuch as very little Soviet military cost data is directly available, it is necessary to synthesize the Soviet costs by estimating the cost-generating characteristics of the Soviet forces, assigning prices to individual elements, and summing the costs of the required numbers. The long and short of this is that intelligence cannot develop the costs from economlc'data- -icalculate the costs of cost factors derived from VS. data,"*
The costing is thus done primarily byeapon system is costcd as if it were produced fn the United States by VS. technology and methods and with VS. personnel. For the sake of comparability the VS. systems closest to those of tbe Soviets arc used and are modified to allow for known differences. Where possible, Soviet factors such as Soviet labor prices fn rubles are used, though this introduces non-homogeneous units, rubles and dollars.
How good are these costs? Since the costing is by analogy with VS. practice, it will be useful to inqmrc how. costing is-
US. Costs
Tbe cost of advanced weapon systems has been increasing rapidly in tbe United States. Dr. Harold Brown bas illustrated this point by comparing the fly-away cost of the World War,er pound, with tint of. to beerhe reasons usually advanced for the increasereatly increased sophistication in weapon system components and the Increased cost of materials and highly skilled labor. It is well known that there are other factors; we shall mention three.
The first Ishe way the business ishe United States, and more particularly In tbe defense industry. Suppose thereesign competitionew weapon system. Two or more companies may cany out very extensive andfforts, often involving the same general technology. It is not unknown for such competitions to become quite protracted, with resubmissions
D. Bnmner, Sooiel At Armaments- and ThttrSanta Monica, California; Tbe Band Corporation,R (Secret, p. 1.
eeting of tbe Iiistitute of Aeronautics and Astronautics in,
required. Each competitor must demonstrate com pet race, in terms of personnel and facilities, to proceed with the contract after it Is awarded. Only one company ultimately wins the contract. Toe loser or losers may be reimbursed under terms of the development contract for part of their expenses, in wbich case this adds toosts. More usually, the company absorbs the loss, erpecting to male it up through profits on other defense contracts, so that the apparent costs of Surxeedmg systxnas-we increased. ^There areway of doing business, but saving nwney is not one of them.
A second factor Is Inefficiency fn BAD. Thereearth of data on this subject, for understandable reasons; companies are not likely to advertise their inefficiencies. But efficiencyikely to beifferent order when limitations on funds require heavy emphasis on economy and reliance on the ingenuity of project leadership andincentives, as against theractices, with cost-plus contacting, emphasis oo massive documentation, detailed control of lower echelons, etc
research byooper1 on the cost of avffian product development disclosed no investigations into relative efficiency among companies of different size, but on tbe basis of an "Introductorybe hypothesized thatarge companies tend to spend substantially more to develop particular products than do smalln Interviews with managers he found that mostarge'spendsoimes as muchmall one toarticularase studymallarge firmrotective coating for similar products, the small one carriedmonth part-time project estimated tohile tbe target's project lastedonths andn direct costs.
Cooper is careful to restrict bis conclusions, avoiding any suggestion that small companies arc more efficient in production activity. But it is the veryosts in tbe VS. m'rrile and space field, rather more than production costs, that haveiew here that heavy expenditures are rec/uired for substantialespecially in the light of competition with the USSR.ctivities are non-standard and difficult to control, thereby leradihg tbemselves to rat-hole expenditures and enterprises.
C Coopers More Efficient in Smallarvard Buttoeu Review,. This article was dravm to my sttcntioo by Tom Clenraa of tbe Rand Corrscrariou.
A third factor wbich pyramids costs is competition for labor aod materials. Tbe missile aod space industry in the United States has over the past few years been its own worst competitor for talent. When projects proliferate, new mvestment in facilities is required.ompetitive economy the pricing system is tbe mccbanism for gaming priority in personnel recruitment, capita] investment, and
fmrrfienxsoni for Analog Costing
Tbe relative tightness of the Soviet economy suggests thats not likely to share the rich man'soney attitude. The Soviet design-team approach lends itself tocommunication and continuity in technology. It has been noted by intelligence that Soviet design goals are usually more modest, less prone to press the state-of-the-art, less likely to incorporate cost-multiplying modifications after series production bas begun, than in VS. practice. Priority seems to be arranged through direct allocation offacilities,than by priceThere are, of course, disadvantages to this way of operating, but It seems economical in trams of costs on high priority programs. All this would suggest that weenuous basis for analog costing.
Before we turn to future weaponather importantof analog costing of current forces needs to be made explicit Costs developed by the analog method depend not only on what is costcd (unit costs) but oo bow much fs costed (systemshe reasonableness of tbe results depends,arge extent, upon the vahdity of the order-of-battleut the objective inwas to validate aod set economic limits on the order of battle. If the vahdity of tbe cost estimates depends upon validity of the order of battle, bow can tbe order of battle be validated by the cost estimates?1
Let us now consider estimating future costs. Tbe question of future technology is immediately raised. There is of course great intrinsic uncertainty in projecting technology mto the future, whether in foreign
D. Bninner. op. cat, p. L
'The ejection that costs to derived may be measured* tpintt eeoooisjc capsbJina svtll be met below.
or domestic, tnQitaiy or dv-ihan arrphcatioo. Charles J. Hitch, DoD, has observed:
he meat important thing to HcderttaodRhe oWmnt roie played byredicates by "experts" of the rendu or nsefulocss ofrojects are highlyalmost si-ays tsle longer tad cost more thaa predicted (by factorsost fad ta wholelod addition ^h> tocbJCgicJhare*oosuasmjj Investment what tt called environmentalabout the kind of new product (bit will be ssSesble orbe unkWn ' environment of the futureill be avaihMe.'
The matter ofod system cost prediction hasonsiderable amount of study because of glaring mistakes in cost estimates associated with VS. weapon system proposals. It has been found that estimates of total system costs made earlyevelopment program may be less than the estimate made when tbe system is ready to be introduced toto the active invent coyatior even higher. Studies Indicaterimary end ovciridtog cause for underestimating VS. weapon system costs has been the tendency to change performance characteristics or the configuration of systems after the cost estimates have been completed.*
In the use of highly uncertain data for purposes of comparative analysis or valuation, it Is desirable that the tmcettatoties be of the same order or otherwise comparable. It may then be possible,ery rough or crude way. to "factor out" such uncertainties on the basis of their corrroarabuity. Perhaps enough has been said to suggest that methods used to costing Soviet military forces may not yield the required comparability. costs. The estimated costs of Soviet forces so derived may thus not express the true cost relationships, either toteriially or with respect to those of VS. forces, implied by the manner to which they are presented to intelligence estimates. This possirnlity may be erJaanced by the problem of rable-doDar cxmversion.
'Charles Hitch. The Character of Research andompetitive Economy (Santa Merita. Calif onus: The Rand3. p. 4.
H.iscussion of VneerUxtnUj to Cost Aaafysk,Rand Onporanon,.tudy ofoDmade b, DoD some time ago revealed that they wens underbid byof
The Ruble-Dollar Problem
Let as suppose thai after some difficulty hi arrivingeliable coif estimateoviet weapon system (and knowing that it ise have in hand such an estimate expressed in dollars (or partly knhe next problem is to convert the do Lars into rubles. Although thereumber of bad ways to do this, there J" ccnm|etely sajbfactory way,.tThere is no singleactor by any method of calculation (except of course tbe Moscow-pegged exchangend all methods of calculation haveThe way this has been done in costing Soviet forces, and indeed the most nearly satisfactory way. is to relate elements of tbe military costs to Soviet economic sectors for which rable-doUar ratios* have been established and to compute ruble costs by use of these
. The same problem hi reverse cannot of course be avoided in inter-preting the economic meaning of the aggregated costs derived through conversion factors. If they are summed and related to costs calculated for previous Soviet defense budgets, we run into the problems we have discussed in costing methodology, costing uncertainty,of how and when costs areeal costs, etc It is quite difficult to draw simple, accurate, and nseful Inferences from comparing such costs, say costs of Soviet general-purpose forces with those of strategic offensive forces or with data oo other economic sectors, not to mention VS. costs of similar forces.
Coet'EffectroerttM Comparison
Finally,erplexing problem as to bow to compare VS. and Soviet forces In terms of costs and effectiveness by any system
of analysis when their relative composition, sophistication ofrelative austerity, and recruirernents for support are so different. Secretary McNamara has inquired "whether tbe Soviet military estab-
f
Aims Strain
lishment has certain espenditure patterns which, as composed to our own, provide more mihtaiy capability for the same cost* u
tend to
3.
Intelligence has long remarked that the Soviets tolerate crude work-maxvship where technical excellence is not requhed. In tbe first Soviet-produced Jet engine the turbine blades were well made but other workmanship was inferior. standards. In Soviet systems the instrum enU ban andhan tn the us eo^jrvalenb;. Then space vd
are believed to be simpler than ours. Because they are largeravoid the costs of miniaturization and associated problemscontrol and.
It is well known that the Soviets have standards of austerity in military forces different from ours. Probably less well understood Is that they also have different needs for rupporting forces and facilities. As the most obvious illustration of this, Soviet general-purpose forces are for the most part deployed in Soviet border areas or Inhe homeland,ubstantial portion of US. general-purpose forces are deployed at great distances from the continental United States. This implies substantial differences in support requirements of all kinds. In short, the Soviet military problem is oot symmetrical with the US. military problem, aod this asymmetry bas implications beyond the costs of differing rnlxes of combat sums.
These twostandards of austerity and Infor support forces andtn some degreeas suggested by Major Ceneral Deanc, the senior US. rnili-tary representative In Russia during much of World War II, in his descriptionrip to the Soviet front after the Battle of Vilna:
Oo the following day we were Erst driven to tbe bendqmiWn of the Filth Army, which was about fifteen mile* west of Vilna. Coksnel Ceneral Erylov was In command and heith hit entire raff. It wasar cry from tbe American conception of an Army beedquBrtcrs. The entire staff consisted of fifteen to twenty cancers who lived and workedew small trailer! scattered through tbe woods. There wu one huge hospital tent, well camouflaged, which reeved bothonference room andeadquarters mesa. Some offices bid ttenographen at work, but most of them did not. We could oof help but think of the enormous in-itaDation and aD tbe office space and facilities found at an American Array headmiarten. It highlighted some very different concepts la our methods of
" Memoraochrm3 to tbe Joint Chiefs.
Of course tbe Russian problem wu euniidejibly different from oon. In tbe matter of supply they had only oxie theater topposed to tbe many all over the world in which we were fitting. Their supply lines wm confined to an east and west rati tod road net, whereat oars eiteodcd beck across the ocean. Toupply deficiency meantfew days' delay, whereas wc had tD wait for the availability of convoys. In the matter of personnel aD Rossia't manpower was close itand ber wuliogjieis to accept losses showed tbe Red anny to rely on sheet
of aa agrarian population already hardened aad for whom the rigors of battle were hole more revere than the rigors of peace. Post Exchanges, failed Service Organisations, doughnut wagons and other morale agencies which call for overhead were unheard of
Tills, from the Russian point of view,uccessful army which had accomplished everything necessary toreat victory. The Russians possibly still carry something of thisheir minds .as they build new military capabilitiesew era. What this means in terms of combat capability has not been tested.
Let us review what we have covered. We know that the Soviets have important resource allocation problems. Military expenditures can be made only at the sacrifice of other desiderata competing for the same resources. But the problem of measuring constraints on such expenditures, we have found, runsumber of conceptual and technical problems having to do with erratic trends In military budgets, inability to derive mission breakdown from Soviet budget figures, methodology of estimating costs, translation of costs into rubles, and forming judgments about them in the framework of the Soviet economy.
Itertain optimism to expect intelligence to be able to cost weapon systems which cannot be described in detailed cost-generating terms, to do it by methods which have proved to be quite uncertain, to arrive at dollar costs and translate these into ruble costsatisfactory methodology, and to estrapolate all this, by any method. Into an uncertainEve or seven or tenarrive at system and force costs which have any useful precision. Or whose precision can be guessed.
After tbe costs have been derived, they must be related towhich servesauge of thehey engender in the economyeasure of economic feasibility. They must be mcas-
"-John Russell Deaae. The Strange Affiance (New York: Viking Pros,.
mod against some such standard ns CNP growth projected into the future, or to tenm of the ucrifice they wouldome other Soviet objective,nvestment for CNP growth. Tbe most rrrrpressive effort to date to assess tbe impact of future Soviet nulitary erpendituxesact adopted the Utter standard. Let us now era mine the attempt to make such assessments in practice.
The ttrtsmeettc*-
Tbe forces postulated In "Intelligence Assuroptioos for Plarming" have been coated and found feasible and reasonable from an eccrnoeakc point of view. Tbe forces listed in "Alternative Ten Tear Projections of Soviet Militaryroup of documents producedIA/ DIA Joint Analysisave been costed and described as feasible. Tbe most impressive and comprehensive effort to date, bow-ever, is the report "Soviet Defense Fapeoditures and Their Economic Impacta" u
Inasmuch as this latter document is tbe first attempt to go beyond simply costing Soviet forces and declaring themt Is irnportant to scrutinize Itsad results. Its ultimateof tbe cost of estimated future Soviet forces is in terms of sacrifice In economic growth. How good is our understanding of economic growth?
Intuitively, there can be no question that competing expenditure programs have an influence on economic growth, and growth on expenditure programs. How much is another matter.4 the mteujgence ccarxmunity. noting that "new extensions of Soviet eeo-DOertic assistance toon-bloc countriesellowil-boo2 and did not0 million" declared, This marked decline cannot, of course, be attributed solelyesource squeeze within the USSR but has almost certainly been reinforced by tbe domestic competition for increasingly scarce resources and by the overall slowdown In Soviet eorxnomic growth... Eecmcxnk aid to
""Hi' artabuatxment of this group seasy W. E. Seidel fa bis article Tnatfligeiie* for Defense PUsIing," Ol.
aW
"No fnraee ethuwise artjaaaiad by tbe aueflienaee dxamaaCy aa feasatteyet bee> deeta/edrUse tons of bitcfiignee cost aWysia.
of the rnetbc^ology are Informifly discussed by George Ediund In hit "Cms or Putter PreMeani of Use ColdXt.
o recover ibe momentum of earner
years."
But5 it had to acknowledge that "tbe hiatus in Soviet estep-sion of econcanic aid lo less developed countries of the Free Weald was ended as new creditsto]0 millionbe rate of erperrditure has been risingnd the dour outlook for the Soviet economyith "chrrmk'ut^ "programs'^
an eraaoomy too large, too enmbersomely Btanaged, and too complex to change gears overnight" became rapidly more cheerful
Another example suggests tbe depth of our understanding of growth In the Sovietnhe New Tori; Timet reported. tbe onceercent annual economic growth rate of the Soviet Union had slipped to LS percent In the last two" This news became available through anCIA statement to the pressuccession of massive grain purchases by tbe Soviet Union. It was greeted with satisfaction by the press butixed reception among VS. and British academic specialists on the Soviet economy.
Toumber of tbe specialists, the statement raid both too much and tooentral problem was the role played by Sovietbe economic downturn. Part of the commentarymore arithmetic than economics. The agricultural sector has been counted asoercent of Soviet CMP, depending on the omission or inclusionind rentith massive crop failures resultingevere depression of soector of the economy, one might oped the CMP to drop severely. Then it might rise even more dramaticallyood crop year.ery bad harvest canne observer remarked,uite moderate harvest can mend. If4 agricultural production (in the Soviet Union] equal* thatnd other things grow as they did this year, except that trade and Ugbt industry expand slightly along with
"Thalugeof China's ecoustnlo growth daring the "Crest Leap Forward- period ii wrD known. See. for (sample, B. B. Rebbcchi, Tcot-Mostrsn: The Chineses SiWxWadllen, "Clsoese Crowtb Estimate*CxSMs VTJ Z.
"Stanley Coha, "Tbe AptcoJture Sector Weight lo an Index of Soviet Cross Nationalhe AS IK fcJUem.ha aaasa sector tn the VS.nly aboutarent.
the increased agricultural supplies, the national income will rise byIt did rise, according to CIA, about six percent. Tbethe Soviet Union is lacedroblem of resourceobscured rather than illuminated by the SL5 percent growthhe general downward trend in the growth of tbe Soviet economy
spacentelligencenote of the enormous costs of our own modem weapon system and space programs, which sharpen our appreciatkin of the economic constraints on the Soviets in similar endeavors Thus the general intelligence htdgroent is that the Soviets are indeed faced with resource limitations andresource allocation choices. How much farther can intelligence go? Let us see bow much farther It has gone in "Soviet Defense Expenditures and Their Economic Impact"
ocrrion
Twoighow, along with the programs they Imply, are oosted on the basis ofuantitative, physical description"in "NIEs and relatedHowever, because many of the official estimates were not expressed in sufficient detail or did not conform to the desired probability criteria. It was necessary toariety of assumptions in order to provide descriptionsfor costing
Wiles, "OAon tbe OA's SUtrsneat ofn the Soviet Eocaaomy" TheEuIielM,
mCf. Alec4 Per Cent and AHoviet Studio,nd Stanley If. Coon. "Comment on4 Per Cent and AH That.'" Sooia Studies,
erhaps the ssngle most Important (actor [contr&ating to the lag to Soviet prodsie*rvity) is the demand of the defense program3 for scarce resources aod highly trainedt ts notewottay that two USIB agencies (one rxm-tn&ary) bave Joinedootnote to the subsequentotter, of NIEand NIEstating that Soviet defense spending uncertsinOex are "too great toudgment at to the general trend of Soviet cfctextte espersditurea particularly to recent yean."
t roust be emphasized that the high and low forces are notat limits or boundsoauWnsa Ileal tense. Yet the synthesized costs ol these forces have been introduced intos tbe range of Soviet military expenditures, as we tee below.
Sorms
PJeitha the rjnrtorni costed dot the variety of assumptions used in costing are described th the report However, the basil for selection of systems is described as foDowi:
If tbahat thereechabOJty ofercent that an (teas wocldas included In both the high aod the low if mi " It however, the probe bfliry wis onlycant that anwould appear, ft was tsscbtded tn the blah aide oafy, aad If tbe probe bully of Its
wai laae taum.SO DaroKht was earafaed been both aides.
bar of teems that wild
be deployed. Thai ascendwas rangedrebeb&ty ofaroene
The costs of the high forces and the low forces were arrived at "by liiripry summarizing the expenditures fear all of the high force 'assumptions' on tbe one hand aod all Use low force 'assumptions' oo the other."
Abo, no attempt was mid* to take esplkit account of unotrtalntta* about tha priceshe costing eaaresse. This decision was governed by piiotl.il considerations, peiUcuUjfy by the desire not to obscure the effects rerulnflj from oacerUlnty a) to physical posrun by nWroducsng raogea that inflect mseereiMsty ei to onct or prion. Thai deeasson Aanll not bel toudenetst that the range of travorrtaiaty at to Soviet com oroSrieody narrow that It can be Ignored.
This, on one page. Is all that is told the consumer about the forces costcl. the methods used, the reliability of the data, or the problems arising outosts In US. dollars and Soviet rubies. Tbe remainder of tbe report is taken up with summarizing *espei>diture implications" of the high and low forces, discussing the 'potential impact of the eapendature series oo tbe Sovietndthe "dollar eouivaients of Soviet defense expenditures."
We shall not try to surrmaartze theages of teat, tables, and charts. Tbe foOowing estract and the charts inre enough for an understanding of the general conclusions.
A possible rate of growth of CNF cooststeot with the high asratsaptkM might averageercent for the whale period; hothe rata for the period7 might be eaaaftaedarcasst perata of growth for CNFercent, however, would be sueeaJeiat with the lowThe empirical basis for thawot eatnuively or rigorously developed as yet, and the Soviet economy may do either ahj-mfir-intly better or worse than projeceed."
hese fodgments were later introduced Intoithout the croalificatioQ in the but sentence.
So vie* A'mi Siiain
USSR: Comporiion of Tola* Defense Etcptndilurttt for Low and Hroh Assumptions of Aiiu'rory forces, by
defense general purpose
1 4 5 6 7 8 9 0
T961 2 IOCS 4 5 6 7 8 9 C
oCvalopMnt. lest, evaluation,pace program
2
The two chirtJ in Figurerom the report to question, show the implied costs of the two programs. The present writer has added the later estimated expenditures to eacht may be noticed that the implied costs of the forces with tbe high and the low assumptions extend backow far, one has no way of determining from the charts. This is as one would expect, because the cost implications of different future force levels do extend backwards m
time-
Thereertain difficulty here, though. The cost implication of the high assumption ft as much asercent higher than the low four years ago and becomesercent higherow the estimated actual expenditures curve falls between the high and low assumptions, which seems reasonable. But logic demands that the estimated actual expenditures embrace both the high and lowin the present and past, as either of these assumed forces could be the actual program at the time, according to tbe arsumptioos by which they were oenstituted. Thus the uncertainty in the estimate of actual expenditures must be at least as large as the difference between the high and low5 percent of the lowf this degree of uncertainty Is accepted In the cstiniated expenditures which have been "straining" the economy fn tbe past and present, what basts is there foruture "strain"?
Nevertheless, wa find these judgments concerning the effects on economic growth of different levels of defense expenoatures appearing inithout the qualifications (which them serves teemed inadequate) that appeared in the study from which they were taken. Moreover, we note data on the absolute magnitude of militaryappearing without appropriate qualification.ives for4ange from9 billionut this range reflects only uncertainty about
"From "Soviet DefenseA/PRbe data are tbe same as those tnt y. perhaps worth noting that tha currant astnaata of Soviet aafiitary esrsendlture* for recent years varies coo-ti'irshly from previous aaConatrs for tha (ansa years.
"Not Id take asts account the furtherbe costtag of the high and low aasaptsoes, or tn tbe Lgh and hrw forces tbeoseW
" These and other data conoernsng raOitary esrrwasdmawppearing usn paref the NIE, are taken from tbe study we have part discussed.
ie forces coifed, not the costing uncertainties with which this essay Is concerned. This fact is not noted in tbe estimate, nor does any sxpression of tbe tenuous nature of this expenditure data appear there It is not surprising that nuscoiicep lions concerning the data arise.
Sumrnany of Uruxrtainties
In discussing tbe means by which mtelligenceuantified expression of Soviet economic hmiffRotxf to produce^^Svanccdwe haveariety of uncertainties. It may be worth while to review them. Military budgets,ompromise among contrary mterests. show erratic treads. The VS. budget has fluctuatedery considerable extent and in an unpredictable pattern. The Soviet budget has also fluctuated, we are quitehow much. Extrapolation does notarranted method of estimating future budgets.
Increments of the Soviet military budget are hidden, so that we have no "pie" to slice into mission forces or elements. The pie must be analytically created by costing assumed elements. Meanmgful costs of modem Soviet weapon systems cannot be derived from Soviet economic data; most costs must be estimated by analogy. costs. The basis for analog costing appears uncertain, and the methods used may produce costs oot representative of the real costs of Soviet forces, especially in the light of incomparable features of tbe respective forces.
Tbe method involves use of both rubles and dollars. These are not horriogeneous units, and conversion from one to the other presents an anomaly which translates to uncertainty. Having arrived at dollar/ruble costs, there is difficulty relating these to some expression of strain or economic hmiration. Intelligence has enjoyed nosuccess in predicting CNP growth. Nevertheless it is in terms of limitations on CNP growth that intelligence has attempted to measure the impact of military programs.
Use in Defense Planning
It is to be assumed that economic intelh^eoce data and judgments contained in National Intelligence Estimates, and costs of Soviet forces synthesized by intelligence at tbe military planners request, are to be used fn defense planning, the purpose for which they wereWe should then not be surprised to find the foUowing reasoning advancedertain sensitive military planning document of considerable importance: The United States and its NATO allies
are spending about tbe same amount on gerier al psirpose forces as are tie Soviet Union and the Warsaw Pacthis fiscal equivalenceasis for judging future necessitiesorce planning point of view. If the spending is the same, tbe essential problem is to see to tbe more effective use of the mihtary resources. Including more effective organisation for employment, rather than adjusting, expenditure tevels^to nuiitary|^rieejs^
Comparative ccooomic measurements involving Soviet and. military equipment and iorccs are appearing wtth ,increasing fre-cruency in military planning documents. One finds statements like "replacement costs (of US. and Soviet equipment) ought to express the relative effectiveness of variousnd charts relating to all kinds of forces with abscissas and ordinate! labeled "Soviet costs* and "US. costs."
tudy which combined and summarized extensive substudies undertaken by the three service* at the request of tbe Department of Defense we find the fouowingescription of fts focus:
If tha Sovietsoflus tn create damage la tha US and theollars to limit damage, what at the percentage US pooulaboo aad loduitijhis oaa be si^eiamf to ttmn* of aaxssaagehe cow fee the US toivenurvtetog" per dollar of Soviet eipervitureIt*
The current trend in tbe Department of Defense seems to be in tbe direction of increasing use of Soviet forces costing and more reliance on economic mtelligence judgments. Indeed, thisatural evolution fn the use of systems analysis for defense planning. An objective of systems analysis is to explore or to refineilitary problem to that marginalerms of scene cost-effectiveness yardstick are Identified and can form tbe basis of conclusions. Marginaloncept familiar to any student of
"The following Matcment ta from aa explanatory footnote la Memorandum to Holders of5 andSUtoel the evidence a> aot adequate for an ritsmaet of bad armaments pro faction {In the So-Set Unioel within uacfnl ranges of confidence, the production figure* usod for coc-puttog expenditures for such proeUtcttao were developed from assumed reouiteuserrU in order to permiT rncluiion of expeoiIor land arrnarrsertti tn the gross total."
Summary Study of Strategic Offensive and Defensive Force* of the US aod USSR.*of tbe Director of Defense Braeaich and.. It should behat that Body we* oWnUaed for Information caly. Nevertheless, data from it have been used as kaput to eefarr weapon ryttem rtudka.
elementary ecoriomics. isrimacy concern of the systems analyst He is interested in the range of oammishing returns, and he seeks to determine where marginal advantages and points ofbetween the cost-effectiveness of alternatives should influence decision making.
It is quite clear that important insights into rnilitary reUtiomhips and the tradeoff nature of various military measures may be derived
clear that"concluV'niay "underInfluenced by inadequacies tn the eccnornic Inputs. Even whenare oot sensitive to such inputs or even related tn them, intuitive comparisons of economic capabilities can scarcely be avoided.they are encouraged by the present circumstance in which data on Soviet military expenditures and costs of categories of Soviet forces appear in intelligence publications at the highest national level often without the slightestart of the problem fa the compartmrotation of the intelligence from the planning functionailure of communication between intelligence and the military planner with respect to tbe adequacy of the data.
But the basicotatter of communication. Nor is the diffkxJty of estimating Soviet defense expenditures and of understanding and measuring eccnoouc limitationsatter of accessibility of data or competence in analysis. Some of ft fa due to the difficulty of the science of economics. Economic theory is not weDnotay to allow transfer of data from one economic frame of reference lo another with rigor, or even to understand fuQy its meaning in one frame of reference. Otherwise how would it be possible for two such distmguished economists as Calbraith and Myrdal to draw opposing inferences concerningproduction In the United States from the same set of data?
Von Neumann and Morgenstern, who have made an extensive effort to expresswKVUe relationships in mathematical form, bave remarked:
o may alto observe thatse reeling of duudUaeUoe with the asatfacsnatscal treatxaaat of economic theory oWrvea largely hnaa tha tact thai Irooumdy one Is oarered not proofs but vtere asserCona wharfs are roaBy ao better than (ha sanaa rati rail green ta hsseary Sena. Very fiwjuentiy the proofsathematical treatment bai been attnnptcd of heldi which are so vast and to complicated thatong time tomooh more esnptrtcal knowledge il aujiujedardly any ieaaoa at aD to eapeet ptogress moreJoha Voo Neamaia aad OnW Mo.grnsMm. TKccy off Feowo-we Bessarawr (hwroo Fitaortan Dar-easaty. p. 8.
In the making of wine, more pressure oo the marc may not improve the product II may be (hat squeezing harder the available Soviet rxtmamtc data, or tbe economic intelligence analyst, will oot achieve the kind of product envisioned by tbe requester. In tact, it may not be too much to say that the pressure bas already been excessive, fudging from tbe product At least it may be time to consider tbe matter.
1 W
Some RecorrrrTieTidmSoTU
If would seem that the expectations of Defense consumers coocern-fag the usefulness of economic mttsugence on the Soviet Union to force planning are quite high and the prospects of satisfying these' erpectatioui quite low. But, far from being informed that theare low. the coorumen are being provided data on costs of Soviet forces to NIEs and other intelligence productsay that can only create nusapprehenuoos roncraming fts precision It may be that some Defense consumers have already been extensively misled concerning tbe basis for intclligence-lupolled data on Soviet military defense expenditures, judging by their statements quoted above and ethers making use of ft. It bas been Included tn weapon lyttem studies in the Pentagon, and there is every mdication that it wiD continue to be used in such studies and accepted at face value.
It would be invidious to imply that those doing the economic intelh geoce analysis do oot understand the limitations of tbe economic figures appearing in finished mtelligexwe. Yet these are not suitably qualified when cited In estimates and studies, and no coherent,statement of their limiutibns basather, when roqusrerraeots for such data are voiced, intelligence uncompbnhingly (and unquabfyingty) seeks to meet them. Why this is so cannot easily be understood outside the frameworkroup of dubiousabout the relationship of intelligence to planriing ensconced fn tbe folklore of tbe bostoess. But If this analysis of the nature and nses of economic mtelbgcnce on Soviet mihtary forces is notawry, It Is clear that the intelligence consumer it ill served by the resulting products.
" ataee-er. there exfatt en .end. ebbenbn* -eefce-ts by srt*fc aach data are derived, ao aa mdrprndent evahuoOn of their preeCsSOB could be Bude. Even tbe robte-doflxr ratios napbryrd have not been peblilhod
For wtelligence, three condiisjoni suggest themselves. First, there appears to be an Important needomprehensive statement about the precision of costing Soviet forces and tbe limitations of economic Intelligence judgments made on the basis of such costing. This could take the formublished study on methodological problems in the production and use of economic mtelligence and mightetailed examination and assessment of the confidence limits of various categoriest data. Tha importance oftudyhat tt might be published under the aegis of the USIB.
Second, there shoulderious reconssdoraHoa of the utility of including costing data In National Estimates and allied documents. This Is not to say that study of tbe problem of derivingmission breakdown of Soviet defense expenditures precise enough to be useful should not be continued. But iferious question whether synthesized data for whicho direct evidence should be certified by Inclusion In National Estimates aod accorded the stature of national inteUigence In planning.
Third, if suchncluded In NIEs, it should be properlyeven if the qualification destroys or greatly reduces Its utility to tbe Defense planner. Possibly it should also be cross-referenced to other papers which more extensively discuss and qualify ft. It shouldule of intelligence that Informs boo be setontext that,inimum, accords the consumer an appreciation of its limits, of what is not meant as well as what Ts meant. There seems torinciple analogous to Murphy's law in industryhich states that if intelligence can be nusintcrprcted, it will be.
Of course these three conclusionsoncomitant study by the planners themselves of tbe uses of eeccaonxic inteUigence in defense planning in the light of problems associated with itsmight be worth while. It might be found that Soviet defense expenditure data received by DoD from the mtclligerice community does not have the character anticipated when it was askedthe utility. It is in the spirit of systems analysis that there should be an evaluation of alternatives in analysis methods as well as In what is analyzed. It seems clear that thereet of potential uses of economic inteUigence la defense planning for which the presently produced data are not satisfactory. There mayet of uses for which such data, produced in the form of assumptions, may prove satisfactory. Io any case, the utility of presently produced ccorsornic intelligence should notatter of presumption.
"Murphy's law:achine can be assembled snoog. It will be.
Original document.
Comment about this article, ask questions, or add new information about this topic: