Created: 8/1/1984

OCR scan of the original document, errors are possible

cla historical review program release as sanitized


offiil PmrJucis.



A Comparison of

Soviet and US

Gross National Products,

A Comparison of

Soviet and US

Gross National Products,

strengthajor determinant of international power. The

ucTihaVC Clcarly rccoenizcd ,his relationship io numerous public

-or um4over the years. The most famous of these were made by

Khrushchev, who boasted that the Soviet Union would outproduce the United States0 and that his country wouldigher standard of living than any capitalist country

Figure I

-.ltd lit.

Soviet GNPercentage of US GNP, Selected Years'

These forecasts have not come true, although the Soviets have made someomparison of the annual outputs (gross national products) of the two countries shows that the Soviets gained on the growing US economy0s their GNP rose fromoercent of the US total.owever, they lost ground, droppingf US GNP1 and remaining at that level

The elements of GNP leading the shift5 were defense and investment. Average annual Soviet growth rates in these areas were markedly lower inh Five-Year Plan than in earlier plan periods and were lower lhan corresponding American rates. The level of the Soviet defense effort, however, has remained substantially above that of the Uniied States since the. The relatively slower Soviet growth in defense and investmentmall relative gain by the Soviets in consumption, even though consumption was by far the largest share of GNP in both countries. While the Soviets gained in both per capita andconsumption, their per capita consumption wa<only slightly more than one-third that of the United States3


CII4 Afiull iti'

We believe the gap between the two economics ttat began to widen5 will continue to increase in the near term. Wc expect Soviet GNP growth for the rest of the decade to average belowercentfor the US economy by US forecasters. If these projections arc accurate, Soviet GNP0 will be back down to less than half the US figure.



Bjcistound and


Trends in Aggregate

Trends in GNP

Implications of tbe


Methodology of Soviet-US Economic

Between Civilian PPP Ratios

Miliury Ruble-DolUr Ratios

Com para lire Soviet and US Defense

of Real Soviet and US GNPs:

A Comparison of Soviet and US Gross National '

R>ckfioood aad Approach

Grots national product (GNP) it lhc total market value of all final goods and services producedation's resourcesear. It is frequently used lo compare the sizeountry'; economy wilh that ofomr>irisons of ihis type are parlkularly important to foreign policy formulation because ihe relative sizeNP represents Iheavailability of national goods and services for panning given policy options. Eccrsomic strength compared with ihai of any other country orof countries isrimary deierminant ofower. As the first US Government siudy comparinc Soviet and US GNPs staled

In tht sirugglt -ihe Communist Bloc ond -hed. Iht factor ofconomic strength is fundamental. Stithernor political stabilityirm miliiary posture can lone be sustained in itsoniinued. dynamic economic growth,or surpassing that of the Communist Bloc, would term to be indispensable for the mainte-nonet of world stability endorce for world peace.1

Periodic comparisons of Soviet and US GNPsserve at yardsticks with which US policymakers

can gauge the relativec strength of the

world's otter supeiTwwcr

' Commanitt Countries do mat measure Iheir cconsmr'lutput inCNPicmu Rather, they iirMii.ni cejncciM of national income thai ticlade* "neKiBroductive" uie) of nwio (meal senieet aad towlA analysts and other Western cononviiiiore awful to aaalyie Soviet economic data in GNP terms tBan acmtdiaf to national income The proccdarc byihe conversioneeesipiithed it ipellod out ia NFAC Research Paper IBUnctauificdlSSR.reantlllailon td Marrtst and Wrtti" Mrosmets of ri'ttomot faremr.C Researchant). C


' Tindi In gfanmlfomnarttm* aftkr Written

Powers and tht Sarin Blot, loin Eccratrric Cornaullcr (IF.Ck Wai hiIS. pp. iii-iv fSeticc of Iranitnlltat from Graver Fmley.iec*or of the Iff introduciaf the ttudy by the Iccisliiarc Reference Scrvicei

The Soviei leadership hat long been keenly aware of the USSR's secondary status at an economic power. Stalin and Molotov both eihoeled Soviet workers to overtake (he West. But Khrushchev transformed the long-held Soviet objective of "catching up" with Americaressing national goal. In thee promised that ihe Soviet Union would surpass the Untied Slates in total and per capita ooipot within the rlecade. In1 he boasted atd Party Congress that the Soviet Union wouldigher living standard than anycountryn addition, he staled that the USSR's industrial output would be twice1 level of the entire rhonsoeialist world inears .

Khrushchev'* dreams did not materialize, but the Soviets have made rjonstderablc progressndustrial production has shown substantial gains, and increased production of consumer durables and food indenerally higher standard of living for the present Sosiet consumer lhan his counterpart two decades ago By no means, however, did (he Soviets "bury ihe enemies of the Revolution" as Khrushchev claimed ihey

Tbe rJeveksprnent of the Soviet economy relative to thai of the United States0 may be traced by comparing trends in aggregate GNP and changes in ihe major GNP caiegories (consumption, inveslmenl, andonsistent comparisons of ihese two dissimilar economies are made by expressing their national outputs in the inorseiary unil of one country or ihe oiher The generally preferred procedure for

ooks. Ne*a hi*hruihchev esiimi thai hii it Heme at about "buryint" ihe United Stale* wa* erUvrcrcteatcel by the WesternHe aid hat mended*ai thai lhcetata- class" would awryevertheless, the Asneeieaa publiche lime pfreewed tht statement >i open economic chillenxe


making international economic comparisons ii toeach country'* GNP into the currency of the other country (Soviet Union in dollars and United Slates in rubies J. Two GNP comparisons,en be made, one in rubles and one in dollars.

The two comparisons will yield different results. This phenomenon is commonly known as the mde* number problem, and il results from differences in the relative prices and quantities of goods and services found in each country. Goods produced in relatively large quantities in either country tend to sell al relatively low prices in that country, and vice vena. Soviet GNP is,arger percentage of US GNP when comparisons are made in dollars, since dollars place greater weight than ruble prices do on investment aod defense goods, which account for larger shares of output in the Soviet Union than in the United States.

The important point about index numbers Is that valuationsubles and those in dollats are equally correctingle figure is needed, economists by convention often resort to the "Beometttc mean" of :he twoeasonable point estimate that falls between the ruble and dollar comparisons This paper will follow the geometric mean convention, although the reader is cautioned thatsed for itsconvemer.ee and does not, strictly speaking,ore valid result than thai calculated in citherhe actual ruble and dollarcan be found in appendix D, but the geometric mean estimate will be given in the body of the paper for simplicity.

Converting Soviet and US GNPs in nationallo ihe other's currency involves tbe use ofconstructed price ratios, called purchasing power parr.ieshese ratios measure the relative values of comparable goods and services found in both countries al the tame time. Price ratios for individual goods and services can be aggregated into collective

'ureinen of indrs aMtn. a* -a* ale inf nana rrf theaa rapt laws, and nnaoes


' Therastr-doi'j'

detente icim'ieiumeshai differentlayed aa

i cinlun PPPonlalArd in amvidtien dicocb bmh irpei of'aliasntncillt called "nbtedollarhis paper <or clarity will arbilC'ly xservt that nomenclature eicluirrely ler Ihe mlllisry

PPPs by using either Soviet or US expenditure weights derived from respective national spending patterns. The resulting weighted ratios measure the relative values of tlcms in groups, such as consumer durables. and eventually in categories, such as"Establisheddministratively set Soviet prices, are used in the raiios because they represent what Soviei consumers would see. "Facior costrices that would be charged if they represented ihe actual costs of resource inpuis, are not used in the ratios because they are not available in sufficient detail. The comparisons in this paper, then, are closer to measures of real iircome differentials than lo measures of relative produclion potential*

Over time, prices in one country change relative io those in another.omparison using the PPP ratio methodology, such as the one made by the CIA, requires periodic updating to remain valid. This paper presents the results of an updatedusing recently completed and revised PPP ratioPP laiiosrices formed the basis of US-Soviet GNP comparisons developed by the CIAn Ihe comparisons published (hai year, Ihe older raiios were replaced by the preliminary results of new resumption and investment PPP ratio studies6he final results of these studies have been incorporated inio this paper.

"Seeuller discussico of the differences beiwee-mea Mies of income differentials aad production potential,he PPP studies upon -Wen Ihe analysis in tan paper i* basedore detailed cap's nation of the elaiiilioiton and

Mint sehemei imol-ed in dcri>in( eoniumption and invni-menr PPPs thai is presenled here. The imereiled reader ii referred

owfipr'on rAr US cud USSR. An ttHinittionol Common*

FACPaper FRSO-IO*IO(Uiwlaui-fiedl.SSR ond ihr Uniitd Stow P'ler Raiitn for7tolloii.ndnd NFAC Research AidUKlauiftedl..

Bublr-Delle'-r article "The Real Grosi

National Prcdoct ol iheSSK Mmvti af

Eronnmir Growth androvided

the Soviei GNP expenditure ucighls and indties used to compare US and Soviei GNPorer time. US cipendiiare weichii and indeics -ere obuined fiom ike Commerceational Income and Product Accounts

' See bilwards. Invnene. el al, "US and USSR: Comrornom olommvtno Tint af'Cfeaafe.omemay be aeeiiiiomed lo iceing analysisore eurreni priceThe dollar GNP comparisons couldeen indcicdore current bue. but ilfell that the charaderiisicncss of ihi price iambic *as beiicr preserved if dollar prices uere left ir. ihr price bisr of Ihe study. Iniej'aiiori of detente, utiirh ii0 rubles, koi judced loempetllnc reason io metiw. ruble prices6C



during the Eighth and Ninth Five-Year Plans (FYPs) were higher than American rates during ihe same periods. Soviet giowth, however, had beensince at least the, and the relationship between the two economies began to shift infavor iohs the decline in Soviet growth rales continued while US growth

Short-term variations in US-USSR GNPgenerally reflected business cycle movements io the US economy. It is not surprisingarket economy experienced more pronounced fluctuationsentrally planned economy. Planning tends to promote economic stability by controlling fluctuations in demand, although plan goals may not be metariety of reasons. Agricultural output, for example, is influenced by weather conditions, and swings inwere the major cause of annual swings in Soviet GNP.ule, though, the Soviet GNP was less volatile than Ihe US GNP. The GNP comparisons moved sharply in the Soviets' favor during ihe US recessions0. when US growth rales dropped well below Soviet rates. Similar shifts did not occur in the US recessions02 because Soviei economic growth had deterioratedoint where it was only slightly belter lhan that in the United States in those years.

GNP Comparisons

Trends In Ag-regale Comparisons

0oviet GNPror^ioa of US GNPercentage points (geometrict rose fromercent of US GNP0eak oferceni5 aad fell back toerceni1' The Soviet Union gained on the United Stales from iheo theecause average annual Soviet growth

Although the Soviet Union gained ground relative to the United Stales overyear periodhole, tbe absolute size of the gap between the twoin goods and services produced annually in-CTcased, whether measured in rubles or dollars3he US recessions2 caused ihe gap io lessen in those years, but the trend has been upward, and the gap widened noticeably1S GNP0 billionillion rubles more lhan Soviet GNV

To jive iheense of the spread between ihe dollar andcomparisons encompassed by ihe r. cornel fie mean. Sovietiiercent of US GNPrn dotlart andercent us rabies. The spread between the Iwo eornpsrnont has fairly steadily decreaacdhen Soviet GNP waserceni of US GNP in dotleri and it peeeent in radio. Ruble prices, -bere imi.beubles; aad dollar prices, as6 dollars ,


Figure 4

Dollar Difference Between Soviet and US


Altotvle 4Mntnct

* Fir

1 rends in GNPefease. The most sinking difference in trendsthe two countries0 has been in defense. The Soviet defense effort passed that of tbc United States in the. Real decreases in US defense spending occurred96 as the United Slates disengaged from Vietnam, while consistent growth pushed the Soviets overercent higher than the United Slates by theinceost-Vietnam

* The eMirniiei in ihir pipe were prepared wftfc diu endinr.reliminary eiiicnitci?3 -ere prepared aiing roufh itcreraie indeiei. There covM not be applied al Ibe convener* leveL bul an ctelilf i< underil will Wend Ihe compjacrii tomoi'iiOM13 '

roweal US defense costs have increased while the growth rate of Soviet defense has slowed, narrowing the defense gap io about one-third"

Defense has also been the category of GNP use where ihe greatest difference in GNP composition between the two economics exists. The United Stales halved Ihe (hare of GNP going to defenseercent in iheerceni

kirjrr 5

Soslelercentage of US'

Figure 6

Soviet and US Detente'

i ii t i i


' oawfvimi mtwt

irn irx

If if-vllt




Trends in total investment growth were also very diffcient. Total Soviet investment grew almost twice as fast as US investment overeriod; ilerceni annually, while US invesiment averagedear. The absolute level of Soviet investment slaved aboutercent of the US level through ihe, rose toerceni in Ihe, and exceeded US inveslmenl1 bygsmal) margin

Even so. the growth rate of Soviet invesiment has been declining since Ihe, and investment's markedly lower growth rates duringh FYP

Figure 7

Soviet Investmentercentage of US*

Flcure 8

Soviet Investmentercentage of US Investment, by'

a major factor behind tbe Uniied Stales' ability to considerably widen the GNP gapS investment, on the other hand, experienced ihe wide fluctuations commonarket economy's business cycles overeriod but grew faster tSan Soviet investment on average.

Soviet gains in comparative levels of investment were most pronounced in tbe area of construe Jon because of the relatively greater facilities expansion in tbe USSR. Soviet construction investment rose fromercent of the US level0ercent1he value of the machinery and equipment component of Soviei investment alsorelative to that of the United States over the entire period bul stayed below the US figure.these trends is the general tendency in the United States lo devote more investment resources to reequipping older facilities and to design moreuse of mechanical and electronic equipment into new facilities.

Th; USSR hasreater share of itsresources to investment than has the United Statesoviet investment steadily inejeasedow ofercent of Soviet GNPighhile US inmimcntndercent of US GNP overeriod" Soviet investment for machinery and equipment (including comparablyteadily increasedercent of Soviet

" US ntuanil inccane accounts miy include tome recrement expenditures thai could be considered investrrvni. men asplant and dam comtruction, in eatctceic* cf government purchase*of fccdi andn adyuitment for this pasiitle aecouBtiiri anomaly has not been made because US national accounu -ere accepted as tivea by ibe Commerce Detriment. Should in adjustment be made, US investment wrnildither share ef US GNP

" This paper uses Ihe standard edition of capital repair as the turn of outlay* on aooeinvent repair of rued asieu. Unlike current maintenance outlays, capilal repair outlays arc not written off at current coils but arc capitalised. The hook vilDe of Ike iisci it raited to reflect thit type nf repair, which It supposed lo eitend the service life of Ihe use)

Figure 9

Composillon of Soviei and US0



0 ioercenihile Sovieiice ion plus comparable capital repair remained betweennd IS perceni of GNP. In the United Slates, machinery investment increasedoerceni of GNP in those years, while USshare steadily decreased fromerceni.

Consumption. The Soviets have gained slightly on the United States in total consumptionoviet consumption over the period fluctuatedow ofercent in tbeigh ofercent of US consumption1

Within the consumption category. Soviet healthshowed the most.dramatic change by steadily dropping fromercent of US health expenditures0 toercent1. Thbonsequence of US health costs rising much more rapidly lhan those in ihe Soviet Union.

Education costs in the Soviet Union began the period at or above the US level, dropped toercent as US expenditures accelerated in thend, but returned to comparable levelsn tbe other consumption categories of food, soft goods, durables, and household services, the Soviets shoved small but consistent relative gains.

lhc Soviet Union is unusual among developingin that its economy has not become increasingly service oriented. The Soviet service sector remained relatively steady atoerceni of Soviei GNP0hile the US service sector increased fromoercent of US GNPJ before falling slightly toercent1. In boih countries, the share of GNP devoted to durables increased, soft goods consumption remained

Figureand slightly improved Soviet consumpiion rcla-

Soviei Consumpiionercentage'oihe United States over the

US Consumption,I?


Pcivrni JO



standards of living are usually measured in terms of consumption per capiu. The Soviet consumer was less well off compared to his American counter-pan when consumption is measureder capita basis than is indicated by the aggregate consumpiion comparison. This is because tbe Soviet peculation wasoercent larger than the American population overeriod. The trends over time of the per capiu and aggregate consumption comparisons, however, arc essentially the same because theof both countries grew at the same averageIear. Soviet per capitarose slightly overeriod, fromercent of US per capita consumption0 toercenis in the aggregate consumption comparison. Soviet health expenditures showed the most change, dropping fromercent of US per capiu health cxpcndiiurcs0 toercenthe other categories of education, food, soft goods, durables, and household services showed slight overall relative gains for the Sovieu.

Ihe'same, and the share going io foodFood consumptionhare of GNP dropped by as muchercenUge points in the Soviet Union andercenUge points in the United States over the period-

Because consumption is by far the largest component of GNP in both countries, changes inajor impact on comparisons of overall GNP. By themselves, these changesended to raise slightly Soviet GNP relative io US GNP during this period, but they were partially offset by the changes in defense and investment discussed earlier. Average annual growth rates of Sovietinh2 percent in rublesercent in dollars) were hither lhan those of the Soviet economyhole. At ihe same time, average annual growih rales for US consumption5 were lower lhan those for ihe US economyhole for that period. These shifts consequently raised the share of consumption in Soviei GNP. lowered it in US

Implications of the Comparisons

Quiic obviously, the Soviet Union did notgoal of CKitperforraingt was. however, slowlyuntil the. Why did it failcatching up during this period, and whatimplications of this failure for future

Soviet GNP growth has beenownward trend since the, but this trend worsened in theumber of reasons. Some were beyond the Sovieu" control, such as bad weather.nternational economic conditions,ecline in the growth rate of the working population. Others included aging of the capitalrequires increasingly larger inveslmenu to keep it

Figure 13

Compostllon of Soviet and US0




mounting shortages of key rawand energy sources. Still others were the results of planning decisions, particularly the6 to Switch from an "extensive"policy that expanded produclion through large increases in capilal and labor loan "intensive" policy of growth achieved by more efficient use ofBad investmeni decisions also led lo insuffj-cienl resources being devoted to transportation, whic'i created shortages of rolling slock and massiveFinally, some of the causes of lhc downturn in growth rales may be endemic to the Soviet system of cenlral planning. The planning piocess, wilh lison meeting produclion quotas, seems lo have stifled innovation and creativity, which are vital lo improving productivity, Lack of wage incentives and limited availability of consumer goods also have been drags on productivity growth.

If ihe US economy continues io perform as well as il has over ihe last year, ihe cap between the US and Ihe

" Tht effect* of lb* Witch in investment poticici it' morein Dt Research Paper SOVJune

I9*J. Ihr Slowdown InSrMn

Soviet economies is likely to widen considerably in lhc nest decade. Opinion is certainly noi unanimous on whether the Untied States can sustain this growth, bul the consensus of estimates developed by leading private forecasting groups shows average annual growth raleserceni or more through theoviet growth, on lhc other hand, is projected by the CIA to be below (hit rale for the rest of thef these projection* prove accurate, Soviei GNP0 will be back down to less than half ihe US figure. There is little reason lo expect Soviet growth lo exceed lhal of ihe United Slates on average during ihe rest of the decade. At best, Soviei GNP0 i? unltkclyorc lhan (he SI perceni of US GNP estimatednd it probably will be less.

- tUtccIeaaxtiaas of averaft aaawaloimiiiafeetcatncial Chase Fcowttmtai fardartiJi. aadercent for the unit prnori esearch rapt. SOVNbiuir*

I9M. USSH timamir fta/nlluMtw loot


Appendix A

General Methodology of Soviet-US Economic Comparisons

appendixeneral description of tbe methodology used in making the Soviet-US economic comparisons presented in this paper. The first section discusses some of the important basic ooncepts of international comparisons. Following sections outline tbe key assumptions upon which the Soviet-USare based and examine tbe composition of each GNP category. Finally, the comparisons are evaluated in terms of their validity, usefulness, and the confidence we have in them. **

Basic Concepts

Companions of any two economies must value each country'( assortment of goods and services in someunit of measure. One of tbe two techniques roost widely used loday to convert data measured in one country's prices into another's price base uses international currency exchange rates; the other uses ptiicbasing power parity ratios. This section discusses (he reasons PPP ratios are preferred, what they measure, and how they are constructed. It alsoa way to handle the sometimes troublesome factwo-country GNP comparison can be made using either country's pricesase.

Piixchasing Power Parities and Weighting Schemes Exchange rates ate determined mainly by supply and demand for internationally traded goods and services and by international capital transactions.they are not satisfactory for ONP comparisons because they do not reflect the relative purchasing power of different currencies over the whole range of goods and services produced and consumed. They are frequently used for comparisons, however, because Ihey are easily available from open sources Tor most

Wc rely instead on PPPbese ratios measure the relative valueomparable set of goods and services found in both countries at the same time when measured in each country's prices. Tbe GNP comparisons in this paper in general are based on PPP

ratios that indicate the number of rubles required to purchase the same quantity of goods and services tbat could have been boughtiven amount of US dollars, and vicehe exception is defense comparisons, whichomewhat different 1

Exchange rates may differ greatly from PPP ratios, especially when comparisons involve economies with substantially different price structures. In fact,have shown lhat the purchasing power of the currency of low-income countries relative to that cf very-higb-income countries is often two lo three times as great as the exchange rate wouldn tbe case of the USSR, moreover, the value of the ruble relative to other currencies is artificially maintained by the Soviet Government and has no directto relative purchasing power. Therefore, the ruble/dollar exchange rate cannot be used asough measure of relative purchasing power.

Comparability of products and services is tbe key to useful PPP ratios. Comparability may be defined many different ways, but CIA PPP ratio studies for the Uniied Stales and the USSR define ii in terms of function and performance. Forhree-phase Soviet AC electric motorkJTowatt capacity that operatespms was judged toype of Soviet general purpose dectric motor usedide variety of industrial applications.

"The piiictuiiaf power ofrubleollar varies widely amonj iirdrWddal eomirodidcf- Foi eitmjJe., a] cent!vcfc canned crsnie juiee la the Uniied Slatesuble boufhl in the USSR, but it look onr IS in the United State* to purchase tbe same amount of frozen cod aold in tbe USSReeiscussion of the differences innd military ruble-dollar ratios. _

. etorld Product and laeomc. Inter national Comaarltoai of Fral Gmi Product (Phaseohns HojAlai University Pros. Baltimore. IvSI, p. J. )

Figure Id

ConslrucHoD aod Use of Machinery aod Equipment PPP Ratio


Croup PPP*

Suc-cneicr, PPP"

ihlici of electric motor oulpul

Type of product

otors Genera ton Ttanilormen Hi(h- vollaje


Value llMKt

of electro-



of mac It Ine rr (roup .

F.kcirrHcehnlcal i:o-erproducts Maifctne loots lorie-presi CalUnt Tooliredion

jiulrumenu Minieii md

rwUI lore-tit

completionnd paperindustry lood mdiiitry PrtrmOf Ifoiii-uaaipon Coniiiueiion Transportation Aulornc-sllei Trietoo nine:

shires of machinery output



Construction Machinerr and couirmienl


comparable US motor with (imilar spcdfrcatioeis and funciioni was identified, and the ruble price of tbe Soviet motor was compared with the dollar price of the US motor toPP ratio for this particular iiem7 .

The usefulnessPP ratiopecific iiem is in its relationarger group, ll is used loPP ratio applicable to an entire product line; that product PPP ratio is used totill larger group, and so oo. Individual PPP ratios are aggregated into more comprehensive PPP ratios by calculating weightedweights reflecting an item's imports rice in total valce of production. Two sets of weighted ratios must be built up. Ratios required to move Soviet GNP from rublesollan are called Soviet weighted because they use as weights Soviet expenditures derived from official

Soviet statistics. US-weighted raiios use UScompiled by the Department of Commerce as

weights tn that US GNP in dollars can be pat Into rubles. .

A simplified schematic of the weighting process for machinery end equipment PPP ratios is shown inbe product PPP ratio for electric motors is circulated from individual motor PPPs weighted by their share of electric motor oulpul. This and other product PPP ratios are then weighted by their share of electroiecbnical group output ioroup PPP ratio Machinery group PPP ratios are then weighied wiih shares of machinery output toPP ratio for ihe machinery and espiipmcni _

Tbc Index Number Problem In maVdnj Soviet-US comparisons,henomenon common to alt international economicso-called index number problem. It arises because tbe comparison in rublesesult different from the comparison in dollars, yet both arc equally legiiimaie

The lackingle quantifiable comparison can be presentationally awkward, so economists byoften resort to the "geometric mean"asisingle comparison. In the case of two countries, ihe geometriche square root of the product, of the two comparisons. However, tbe theoretical justification of the geometric meanalid average indicator in economicisputed,Becker argues against its use in. US and Soviei comparisons because "the average reflects neither the dollar nor the ruble price pattern and in principle,implyaul Samuebon. on the other hand, believes that use of the geometric mean "docs seem morehb paper adopts the geometric mean convention, but the reader should remembereometric mean represents neitber the ruble nor dollar comparison and may therefore be inferior to the more general estimate defined by them.

Key Assumptions

For ihe comparisons in this paper lo beumber of analytical conditions must be met or al least reasonably approximated. First, the base year for lhc PPP ratios must beeasonably typical year from an economic standpoint. Thein this study are basedatopic of PPP ratios6 prices. The comparisons, therefore, are onlylive of other years to tbc degree that economic conditions6 and. in particular,prices bothiven country and between conn tries can be assumed to be typical for ihose other years. Unusual supply or demand eondiiions in cither country6 would distort Ihe relative prices and cause ihe comparison* to be biased

"Coropjrisoni of Hinted Suit* and USSR NationalSome Rules ol iheorld feJiMrr. vol XVIII. No. I.

"'Analytic Note*on InicrMiMaslornero-iwmfe Journal,1

The choice6ase year hss merit for several reasons. On the Soviet side,6 decision to reduce the rate of investment growth took time to work its way through tbe economy. Its most dramatic effects on industrial production, consumer prices, and GNP seem to have been delayed. as shown in table I.ay have been atypical,6 was more or less an "average" year by Soviet standards.

For 'he United6ear of strong economic growth following the severe recession ofIn thb sense,ypical of the growth rates in.eriods, ll also was not marked by an unusual swing in industrial production, and inflation had not yet reached the high rates of the.

Comparisons for years other than6 base year involve greater uncertainty the more distant thearc from the base year. Changes in relative prices from the base year relationships arc more likely. In addition, US GNP in current prices must be deflated6 dollars using price deflators that arc subject to error. Soviet GNP must be estimated for years other6 by6 weights with quantity indexes of output, and thb process inevitably introduces some distortion. Comparisons of US and Soviet GNPs0or example, are therefore less reliable than those

In thb paper we assume thai valid comparisons of defense efforts can be made using ruble pricesear other lhan the one on which the PPP ratio studies were based. Thbecessary because we make our estimates of Soviet defense spending in0 rubles, and we have no reliable basis for moving these data io an6 base. Ruble estimates of GNP in thb paper, therefore, were also made in0 rubles rather than6 rubles to better facilitate the incorporation of our defense estimates. Thbindexing the ruble prices in ihe PPP ratios0 prices. The reindcxing was carried out at the

Table I

US and Soviet EcoBook lirdiatoti

level of aggregation possible, usually ihelevel bul .down Io Ihe group level in some cases. Wc recosniie that this process could introduce some distortions in the comparisons to the extent0 ruble price relatives are different from ihoseC decided this would introduce less distortion, however, than irbitrarilyate of inflation for defense.

Lastly, wc assume that reasonable comparisons can be made for dissimilar economics. Comparisons arcconsidered more reliable for countries with similar levels of development and social sysiems. For countries at Jifferen* -lagcs of deveropmeni. such as the United States and tbe Soviet Union, comparisons in alternative prices tend to overstate the relative position of the less advanced country. The more advanced country could shiftess complicated product mix more easily than the less advanced country could produce the more explicatedof goods and services of the advance! country. Tbe Uniied States turns out goods ond services that the Soviet Union cannot produce at all because of the widely noled Soviei difTrcully in manufacturing high-quality and technologically complex goods. Therefore.

the ruble comparuon of Soviet and US CNPsoverstates the relative abiliiy of the USSR to produce the US output mix because ii understates the ruble costs of US state-of-the-art technology. The dollar comparison understates ihe US abiliiy tothe Soviet output mix because it overstates the dollar costs of advanced Soviet goods.

Coverage of Product Samples and Comparison Categories

GNP in this paper has been divided into five end use categories: consumption, investment, defense,and other. The composition of each of these major categories and their ratios will be briefly examined in this section."

For| avore detailed description, of tie -ort provides thepaper, see NFAC Research Paper PRO

Jhe UntiedRiiotfor7otlon. voti. INFAC Research A* ER

errw/orw-prira,te USUSSRnd USSR: Afeaswr. of CconvnlendI)


Consumpiion includes all househoTd expenditures on goods and services plus government expenditures on health and education. In the USSR the state is responsible for practically all expenditures forwhile in the United States householdsignificant share. Consumption is divided into six subcategories: Toed, soft goods, durables, household services, health, and education.

The food estimate is basedample of pricestems, such as milk, ground beef, white flour, eggs, coffee, and carrots. Ruble prices primarily represent weighted averages for food sold in Soviet stale retail stores and on collective farm markets. Food produced and consumed on farms is alsovalued at average prices received by farmers for urban marketings, Tbe dollar prices consist ofweighted average prices, including sales tax; for the same set of food items in the United States.

The price samples for soft goods (such as clothing) and consumer durables (such as color television sets)tems. Retail established prices were used and no attempt was made to account for black-market sales, which are prevalent in (he Sovietonsiderable effort was made to account for quality differences between Soviet and US goods, which frequently resultedetermination that the price of the US good should be lowered.

Comparisons of household services were made fortems covering housing, utilities, publicpersonal communications, personal care, repair, automobile services (gas, oil, and maintenance),and miscellaneousPPs forservices arc based on national average prices for individual services such as monthly telephone service and charges for hotel rooms. Although in the USSR most of these services are provided through public organizations, prices of privately supplied servicesn included in the PPP ratios where

The weak link in ihe household services comparison is housing. There is no Soviei counterpartingle-family housing, which comprises (he bulb of hcosing in (be United Stiles. Tbe PPP ratio for housing used in (his paper is (be weighted sum of ratios for rental costs per square meter and for maintenance. The PPP ratio for housing is based on the price per square meter of average-size urban apartments in the USSR adjusted to exclude the latge Soviet subsidieshe comparable US average rental rate is derived from the relationship between US rents and various housing characteristics presented in United Nations-sponsored international comparisons. Ratios for maintenance are based on prices of building materials such as plywood. Portland cement, and paini. along wiih an estimate of labor charges

PPP ratios for health and education are based onaad current materialo tangible measure of output of tbese services exists lhat can be pricedomparison The drawback of the input method is thai il ignores capital inputs and implies thai (here is equal productivity in US and Soviet health and education ororrection can be made for unequal produetivi'ty. The input method also assumes equil quality of service. Most observers agree that Soviei health and education are inferior both in productivity and quality to their USbui there is no ob/eciive way lo correct for these dilTerences. I'c; ihese reasons, the health andcomoarisons probably arc biased in the Soviets' favor.

an ban ticlodod bcoaaae taey arc aunpro-dormc transaction) GNP itiemau aa eaptare oal* iiamaeUao

Oil axitrlbvM leant'-lortadii sabod'O

mrstau aaaaal avian

"Charga la> fiiM ca>U. Mmeea ere tadaded oa talk uGa ma

tkt cautery, la tW I Snin iht Maria* (ouaof

upboUCiial hcasaul atatBantai vftod tor diafimii and irutmcc!

ajor uu itraliki PrcacM oejuMnthaaMafihealth eare beonc PPP ratios lor ratd<al aQnpmcnt areeertaia to be hither mm thate leerncxi! tuppliri

Cor (he United States, miscellaneous services cover mainly finan cial service) that hai no counterpart in tbe USSR. The mivella-neous category was convertedPP ratio bawl on that for personal care aad repair services



Comparisons are made for gross fiiod investment in machinery and equipmeni (includinc capilal repair) and construction. Investment comparisons exclude inventory change and net additionsrvesiock herds,e included

PPP ratio* for machinery end equipment are lakentudy lhal compares US and Soviet machinery pricesampletems. For many years, comprehensive data on prices and specifies lions for Soviet machinery items were not available fromsources. Soviei publicationarge number of price handbooks (tsenjukl) containing priceside spectrum of machineryystematic comparison of US and Soviet machinery possible. The ciassincaiion system of the Soviet input-output table was used to provide the framework for the price sample. Products werefrom theectors that define Soviet civilian machinery production and matchedontractS manufacturer wilh comparable USThe tttnnlki ruble price and the dollar price formed the item PPP ratio.

The PPP ratios for consinjctlon are basedampleypes, of projectsroad cross section of const ruction. Tbc individual ratios were grouped into eight conslniction categories to mesh with categories used in each country's published data on construction expenditures. Construction ralids rely on Iheapilal census harrfrjooks. which provided simplified formulas for cost per square meter of construction of various sires andhese ruble costs were adjusted upward byercent io allow for Soviet cost overruns. m

The quality of Soviet construction has often been observed to be inferior to American. Allhough many aspects of this inferiority relate only io appearance, others reflect the Soviet attitude toward maintenance As an approach to handling the quality problems, we arranged ibc comparisons so that the highest quality Soviet conslnictioniven type of project was matched with average and poor quality USThi* procedure is arbitrary aad does not entirely eliminate the problem, -'though Ihe correction moves in the proper

Capital repair expenditures are ilic sum of Soviet spending on noncurrent repair of fixed asscis. Unlike current maintenance expenditures, capital repairare not written off as current costs but are capiuliicd. The book value of the asset is raised to reflect ihis type of repair, which is supposed to extend the service life of the assci,

Half of Soviei expenditures on capilal repair have been excluded in US-USSR comparisons of new fixed inveslmenl. This compromise is adopted becauseexpenditures are rarely capitalized in the United States for tax reasons, while in the USSR capital repair is an accepted alternative to new investment and represents an increment in tbe value of fixed capital.

Soviet capital repair was allocated to machinery and construction according lo CIA estimates of itsdistribution. The machinery and construction ratios were then adjusted to reflect iu inclusion.

Defense, Administration, and Other Expenditures Defense expenditures in both countries includeon military equipmeni. construction, personnel {excluding transfer payments in tbe form ofperations, maintenance, military RDT&E, andatomic energy programs. Except where otherwise noted, defense will be defined according to US conventions. This is done to make the comparison of US and Soviet defense activities consistent,and familiar io US policymakers. There is some evidence lhal the Soviet definition of defensein Soviets broader than* the US definition. The cost differences between tbc Soviet "broad" and the US "narrow"costs of civilian US space programs that would be funded by the Soviei military and civil defensebeen arbitrarily allocated to tbe Soviets' "other" category.uller discussion of tbe reasons two different definitions of defense are

" Coraprcaemive official Soviei data on defense ipendini have nevee been released. The official fisun:nnually obviously oaversmall part of the loul effort.




Alternate Measures of USending in Selected Veers

and state and local government employees not cm-ployed in education. Ruble wage data are the average annual salaries of employees in Soviet administration categories listed above.

J 1 O 1

2 9 9T3

GNP J 1 1

This study uses US defense expenditures and price dcflaiors from US GNP accounts for consistency with other dements of US GNP. The figures in this paper for US defense expenditures6 prices differ slightly from deflated defense expenditures derived from Ihe US budget and ihe Five-Year Defense Program, which arc used by tbe CIA In other studies for comparisons of US and Soviet military activities. Defense expenditures in US GNP accounts are the sum of actual disbursementsalendar year as reported by the US Department of the Treasury. In contrast, tbe defense series used by tbe CIA is derived fromutlays reported by DoD in tbe budget and deflated by specially construclcd deflator indexes. Most of ihe small differences arise from differences between the price deflators associated with the two series. Both estimates of US defense spending are shown in tabic 2

For tbe USSR,he sum ofexpenditures on general agricultural programs, tbe fbiest economy, state administrative bodies at all levels, and municipal and related services (includingor tbe United States, this categoryesidual of current government spending on goods and services not classified elsewhere. Some of the large: categories of administrator, expenditures are central administration and management, public safety,resources, and agricullure. The totals for tbe two countries may not be entirely compaiable, and Soviei expenditures may be understated. The USSR issecretive about such mailers asfires for police and fire protection, for

PPP ratios for administration arc based on inputs of manpower and materials. Dollar wage data arc based in the average salary of federal civilian employees

For the USSR, other expenditures include thebetween total CNR and the sum of consumption, investment, ccfer.se. and administration as well as several accounts that have been transferred from other categories for compatibility purposes. Oihcr expenditures in Soviet GNP consist of net exports, inventory change, civilian space, civilian research and developmenl, net change in livestock inventories, and any statistical discrepancy. Livestock inventories and civilian research and dcvclcipment funds have been transferred to other expenditures from investment, and civilian space has been transferred from defense. For the United States, other expenditures include net exports, inventory change, civilian space, foreignasstsuncc. and any statistical discrepancy.

Alternative loirrpretalions and Major Uncertainties

Because of lhc wide gulf separating ibe processes of price determination in the United States and lhc USSR, there are two general ways to view tbe comparisons' as measures of relative productionand as measures of relative real income. The first requires some major assumptions given theof tbe available information, while the second is iubjeet lo strong qualificalion.

To measure production potential, prices in the two countries in the base year should satisfy theconditions governing producer equilibrium: they must be free loo that relative prices can reflect ibc marginal rales ef substitution. For the United States and especially for the USSR, the established prices used in ihe comparisons involve distortions of Ihese coodiiions. In the USSR, indirect taxes, which Tall almost entirely (although unevenly) onand an enormous bill for subsidies lo agriculture ind other sectors ensure Ihat relative factor costs will

.be quite different from relative established prices. At the same time, investment goods arc effectivelyaad distributed at prices that probably do not fully recover production costs. XJolU detailedof the established prices to factor costs can be made, GNP comparisons will not provide goodof relative production potential.

The comparisons are more valid as measures of real income duTcreoiials. To reach this tolerpreution. we must assume, as in all such comparisons, that the results do not depend on the distribution of income within the countries, aod that Soviet and American tastes arc tbe same.'* Once tbe proposition of common tastes is accepted, the comparison of real incomes requires that relative prices be proportional to tbe relative marginal utilities of the goods and services as judged by the representative consumer in each of the two countries. Even for the United States, where relatively free markets prevail, there are important distortions from these conditions, in the Soviet Union the prevalence of queues and black markets shows that at state prices tbe Soviet consumer would like to spend more than be or sbe is able to on some commodities. Thus, relative prices are notto marginal Otiiities of goods and servicesubsantial part of the Soviet market. Relative prices apparently also playmall role in Soviet investment and defease decisions. But the assumption that the population adjusts its purchases so as to maximize its satisfaction with given incomes makes sense for both countries.

If the comparisons qualify as reasonable measures of tbe relative size of real incomes in the USSR and the United States, wc believe they trace the upper bound of the ratios of Soviet-to-US GNP. The GNP ratios tend to overstate tbe Soviet relative position because:

The PPP ratios for machinery, equipment, and construction were not adjusted to account for US quality advantages beyond those reflected in the original product matches.

stablished prices in the two countries ignore the substantial advantage (hat the American consumer has in terms of convenience, variety, and availability. These "services" arc covered in the US price but not in the Soviet counterpart. Tbe dollar value of Soviet output, therefore, is overiUled and the ruble value of US output is understated.

PPP ratios forhealth aodare too high because they do not adjust adequately for the higher qualifications of American workers in health and education.

On balance, then, Soviet economic performancecompares somewhat less favorably with that of tbe United States than our calculations suggest

In the area of basic data quality, we have the most confidence In the US- aod Soviet-weighted investment and consumption PPP ratios because of the extensive research underlying them. The US-weighled defense ratiO (tbe one used to convert US defense from dollars to rubles) has tbe least confidence because of the difficult nature of pricing US weapons in rubles. Wc therefore realize our estimates of US defense in rubles are subject to considerable uoceruinty. Thein (he US ruble defense estimate, however,imited effect on the aggregate comparisons. The US-weighted ruble-dollar defense ratio would have to be in error by at leastercent to shift Soviet GNPercentage of US GNP by as muchercentage points, and we believe it is unlikely to be offis much. Errors in Ihe US-weighted defense ratio would have no effect on the aggregatein dollars, since they are determined by Soviet-weighted ratios in which we have higher confidence.

See Kravis. op.efend of the analetous(oe eevntriea aa different as Kenya and Ihe United Siatea

Appendix B

Terencesliin PPP Ratios ond Military Ruble-Dollar Ratios

The incorporation of defense Into Soviet-US ecoriomic comparisons requires combining defense comparisons using the military ruble-dollar ratio methodology and civilian comparisons using tbe FPP ratio methodologyoherent structure. Consequently, tbcbetween the two methodologies take on greater significance than when civilian and militaryare estimated separately using their respective methods. This appendix examines these diffcrerxes and addresses issues of compatibility.

Tbe CIA's purpose in estimating dollar costs of procuring and maintaining Soviet military equipment is to determine what il would cost to manufacture and maintain ihe Soviet design (cielusive of RDTeSF. costs) in the United Stales using US manufacturing technology and prgaices. We do this, when possible, Ihiough deMited engineeringhere we do noi bave enough information for an engineering study, ihe dollar costs of Soviet weapons derived using ihe military ruble-dollar ratio melboclology often depend on mathematical models calledhese calculate costs based on performance parameters. For example. US aerospace industries have found from experience that factors like weight and speed are good predictors of aircraftER used to derive dollar costs for Soviet aircraft, then, would be based on US manufacturing and design exrxrience but would have Soviet aircraft performance characteristics as inputs. To avoid over-Hating the dollar cost of the SovietS CERevel of technologypriatc to that of the Soviet system is selected If no CER or directisS analogue thai matches ibe Soviet good ns closely as possible ii usedast reson. Because of Ihe problems in matching, this method is only usedimited number of cases

Tkt rrtvtu ef Uk atatiury tWIirntdUfcncc Aisewsnnrl SOV

lnd US Odour AeiMUti.


producing then the United Slates. In reality, however, the two ratiosuld not be equal, because it is very eSifficuli to find perfect US analogues. Selected parameters may '< matched, but US military goods tend to be produced with expensive "state-of-the-art" technology that improves iheirin other areas relative to the Soviet good. The dollar cost of tbe US analogue would thus be higher than the dollar cost of tbe Soviet good derived from the military ruble-dollar ratio methodology, making the military PPP ratio lower lhan the military ruble-dollar ratio. This relationship may not be true in every case, but is probably true in the preponderance of cases because of tbe US lead in many keyareas.

disadvantage in producing complex poods, re flee led in high military ruble-dollar item raiios, is beingas the Soviet weapon mix begins to include more of such weapons, Thcic complex goods, however, would have somewhat lower ratios if the PPP concept were used for (he military, which would resultarger estimate of (be Soviet defense effort in dollarsmaller estimate of the US defense effort in rubles. Thus, the use of military ruble-dollar ratios tends lo offset somewhat the uncertainty in ihe Soviet-US comparisons that comes from ihe index number problem in economic measurement

differences between PPP ratios for military goods and services and actual military ruble-dollar ratios suggest that the use of the latter probably affects the GNP comparisons somewhat, but the effects are difficult tose of military ruble-dollar raiiosNP rtudy primarily based on PPP ratios probably understates the site of the Soviet defense effort in dollars and overstates the siac of the US defense effort in rubles relative to what would have been estimated had PPP ratios been used for military goods and services. This is the opposite direction of the bias introduced into Soviet-US comparisons by the index number effect. Tbe Soviet comparative

" Empirical efforts lo estimate these differences basedimited sample, in fact, contradict the hypothesis lhai military ruble-dollar ratios should be hither than PPP ratios for the military, because> types of ratio* are fairly consistent for some goodi Certain durable goods, such is transport aircraft, tractors, trucks, ponitruc-tion equipment, electronics, and traaiport iSips. ate common lo both thend military samples of good) aad services used in oer defense and GNP comparisons. The PPP raiios and the militaryatios for these foods were derived Independently, but the civilian PPP ratio* may be viewed as "military" PPP ratios because ihey relate to goods in the military sample. When(be military ruble-dollar ratios of these goods fallomfortable range of the corresponding civilian (conjeaaralPPP) ratios. When tbe meant of the latlos in each category are ranted, the same pattern '* found in tolfc sets: th; irreee ieehro'.oii-cally complex the category, ibe higher iu mean ratio. This indicates lhat It it relatively more difficult (measured ia teems of eosi) for (he Soviet Unionuild complei goods than it is for ihe Uaitcd Slates. This evidence is suggestiveeneral trend bul not eooduuic because the number of items common lo both seta of ratios and their impact on cither thecomparaiiVeCNPestiinsieor ihe estimate of comparative dollar costs of defense activities is small. Also, most of these (Kds are not as technologically eompki as some military systems, and ihe practicalwrcn Ihe ratio concepts may be larger for more complca itemi

Appendix C

Measuring Comparalivc US and Soviei Defense Burdens

Defense burden is conventionally expressed at the ratio of the value of military goods and services produced to the total value of goods and services produced by an economy when both defense and GNP are measured in indigenous currency terms. The burden ratio answers significant questionsountry's resource commitment to its military, but these answers differ depending upon the definition of defense used."

efinitions of Defense Costs

When we compare US and Soviet defense activities, we usually* dcfmilion of defense because it is more familiar to US oolKymakers. This definition

National security pitoth countries that in the United Stalesbe funded by iheof Defense

Defense-related nuclear programs in both countries that in the United Staid would be funded by the Department of Energy

US Selective Servicend their Soviet counterparts.

The defense-related activities of the Soviet Border Guards and the US Coast Guard (u)

Soviet defense costs measured by this definition es-ceeded US covisonsidciaMc marginyercent in ruble Icims ant' byeicent in dollars. .

A comparison could, of course, aho be madeoviet dcfmilion of defense Weoviet definition would be broader than ihe US definition

- TV readerw- luHiii l> icforrd

S ipeisrwi wmm in The

Bs'dta ti Sokiint*n'atf

pnflii'fl'ii'bV'1 CDA.



Costs of US and Soviet Defense Activities!

Bilfiom6 Dalian

Table 4

Ratios of Defease Costs to GNP fo* ibe USSR and tbe United1



Tbe Defense Burden Ratio

Narrow and broad derations of defense, along with comisaralive GNP data, allow us to compute aof different measures of defense in relation to national output. When comparing burden ratios for differentommon definition of defense sboold be adopted so as not to distort the meaning of comparative differences. When one particularperception of its defense burden is of interest, however, thai country's own definition of defense or an approximation of il should be used.

Regardiess of the definition of defense activities used, any burden analysis should be calculated incur rentes Price and ejaantily rclationsHips vary from country to country, and measuring burdens in indigenous currencies preserves ihe domestic price and quantity relationships that would be distorted if foreign prices were used.

hows the estimated burden ratios based on boih the narrow and broad definitions of defense. Of the eight ratios lhal can be calculated, only ihose based on dollars for the United slates and rubles for the Soviet Union embody all of Ihe rsecessary pricing conditions io serve as burden measures. If ihe Soviet burden of defense were being compared lo theil would be over twoalf times aserceni of Soviet GNPercent of US GNP using tbe US definition of defensehe burden of defense in each country, however, would9 percent of Soviei GNPercent of US GNP. where each figure is basedefini-ilon of activities mosi consistent wjih national ,

Original document.

Comment about this article or add new information about this topic: