INTELLIGENCE REPORT
US AND USSR: GOT^PARISQM OF SIZE AND USE OF GROSS NATIONAL
CIA HISTORICAL REVIEWELEASE AS SANITIZED
DIRECTORATE OF INTELLIGENCE
Psfie
Summary
I. Introduction
General Considerations in International
Comparisons of Groso National Product
Index Number Problem
Meaning of the Two Comparisons and of the
Geometric Average
II. Method of Constructing the
at lea tion of Gross National Product and
End
of Ruble/Dollar Ratios
Calculation of the basic
Adjustment of Certain
and Biusos in the Data
for Raising Individual
Ruble/Dollar Ratios for Grose National
Product and End
III. Comparisons of Relative Slie or US and Soviet Gross
National Product and Major End
A. Comparative Size of Total Gross National Product . Coaparative Size of Major fcrd Uses of Gross
National
Size of National Policy Expenditures .
of Present and Previous Estimates
of US and Soviet Gross National Product
Appendixes
Appendix A. Derivation of Basic Ruble/Dollar Ratios
Appendix 3. Statistical Tables
Appendix C. Source References
- Ill -
Gross Hational Productercentage
of US Gross National Product,
and USSR: Average Annual Bates or Growth
of Gross National Product, by End Use,
Hational Policy Expendituresercentage
of US Hational Policy
k. Comparison of Present and Previous Estiraatcc
of Soviet Gross National Productercentage
of US Gross National3
Derivation of Gross National Product, by End
Use, in Current
Derivation of Major Subcomponent0 or Conoumptlon
and Investment In Current Dollars,
Gross National Product, by End Use,3
6. USSR: Gross National Product, by End
5 Rubles,
9. US and USSR: Derivation5ollur
.0. US: Gross Halionul Product, by End Use, 5
Rubles,
il. USSR: Gross National Product, by End Use,
3 Dollars,
.2. US and USSR: 5ollar Ratios for Gross
National Product, by End Use,
3- US and USSR: National Policy
M. Previous Ruble and Dollar Estimates of Soviet and
US Gross National Product, by End Use,
- Iv -
Following Page
Cfi-'ir-.i;
Figure 1. US Md USSR: Gross NationalEndU
Figure ?. US and USSR: Gross National
3. US and USSR: Major End Uses
of Gross National
k
US and USSR: National Policy Expenditures
nd
US AND USSR: COMPARISON'S OF SIZE AND USK
OF GROSS NATIONAL PRODUCT* "
Summa:
This report present* new estimates of the size of the Soviet economy compared with that of the United States. The newshow the total gross national product (GNP) of the USSR and each major end use of GNP except defense to be significantly smaller than indicated by previous comparisons. The relative size of Soviet GNP is lowered by aboutercent, but that of investment and government administration is smaller by aboutercent. The difference is partly the resultowerof tho ruble value of Soviet investment but. moreof an arbitrary adjustment of the ruble/dollar price ratios for certain components of GNP. These adjustments attempt to take intohe lower quality of Sovietgoods, which in previous comparisons was assumed to be the same as in the United States,he lower productivity of Soviet workers in health and education and in administration, for which no allowance was previously made. These adjustments correct, in part at least, for the most important known biases in the US-USSR comparisons. The new comparisons arepreliminary and imprecise and arc believed still tothe sire of the Soviet economy relative to that of the Uniled States.
The new estimates show that4 the gross national product of the USSRas considerably less than half of that of Ihe United States. Consumption of the Soviet population compared less favorably than GNPell below two-fifths of US consumption. er capita basis, Soviet GNP wasercent of the US level, and Soviet consumption wasercent of that of the United States. But Soviet investment.
* The estimates and conclusions in this report represent the best judgment of this Office as* 3 dollars. Based on lhe geometric average comparison.
little over two-thirds of US investment, and Soviet defense expenditures were somewhat more than four-fifths of those of the United States. These relationshipsepicted in- did not change significantly
The relative positions of the USSR and the United States with respect to total GNP and each of the major ond uses shifted significantly5he most striking changes having occurred in the first four years of the period. 5oviet GNP rose fromercent toercent of the US level, about where it stood also Soviet consumption increased slowlyercent of that of the United Statesromercent5 toercent4ost of the gain having taken place In contrast, dramatic changes occurred during the period in the relative positions of investment and defense. The size of Soviet investment relative to that of the United States rose spectacularly- fromercent of the US levelithmall further gain having been achieved The relative size of Soviet defense expenditures, on the other hand, decreased markedly compared with those of the United States. oviet defense expenditures exceeded US expenditures byercent, whereas8 they had dropped toercent of the US level, and4 they were only aboutercent of the US level. oviet expenditures for government administration were aboutercent of those of the United States, compared wilh
ercent
ponents of GNP that may be termed national policy They include education, industrial investment, civilian research and development, foreign aid, and defense. 3 they representedercent of total GNP (in dollars) in the USSR andercent in the United States. The first three components reflect in broad terms the relative commitment of the two countries to the promotion of economic growth, and the latter two items reflect the commitment to the furtherance of foreign policy objectives. Takenhole, Soviet national policy expenditures
US and USSR: Gross National Product, by End4
(measured in dollars) wereercent of those of the Unitedmall gain from the position Expenditures connected with economic growth (geometric averageowever, rose much more rapidly in the USSR than in the United States. Thus Soviet outlays on education wereercent of those of the United States5 andercentndustrial investment rose fromer-centercent of the US level; and expenditures for civilian research and development rose fromercent toercent of the US level. With respect to foreign policy expenditures, Soviet outlays on foreign aid increasedcompared with the United Statesercent toercent measured inut Soviet defense expenditures declined substantiallyercent of those of the United States.
The comparisons shown inre the geometric averages of two comparisons of total GNP and its end uses in the United States and the USSRone when the GNP's of both countries are valued in dollar prices and the other when both GNP's are valued in ruble prices. The method of calculating the comparisons makes use of the domestic purchasing power of both the ruble and the dollar for specific goods. These ruble/dollar price ratiosepresentative sample of goods and services in each economy are used to convert the components of Soviet GNP into dollars and, conversely, the components of US GNP into rubles.
The two comparisons, of GNP, the one in rubles and the other in dollars, are substantially different. The ratio of Soviet to US GNP4 in dollars wasercent and in rubles, ercent; for consumption and investment the ratios werendercent in dollars andn rubles. These differences in the ratios are the result of wide differences in the patterns of output and in the relative prices in the two countries. Since eachs prices reflect its own and not the other country's costs and scarcities, neither comparison is preferable to the other. The geometric average of the two comparisons is believed toetter measure of the relative production capabilities of the two economies than is given by either the ruble or the dollar comparison.
I. Introduction
Comparative gross national products and their distribution by major end uses (consuaption, investment, defense, and administration) ln the United States and the USSR are used extensivelyariety of purposes The ccnp&risons of total GNP's provide the most comprehensive available measure of the relative sizes of the US and Soviet economies and of their relative performance and potential. omparison of the relative sizes of the major end-use components provides some insight into the policy choices of the tvo countries in allocating resources to the welfare of consumers, tbe future growth of the economy, aad the strategic objectives reflected in military and space programs. Finally,measures of production are used selectively by both sides as propaganda in the cold wara natural result of the great emphasis placed on economic competition between the United States and tne USSR, not only by the Soviet leadership, with its boasts of catching and surpassing the United States, but also by the rest of the world.
This report presents new estimates of the relative levels and rates of growth of GNP and Its major end uses in the United States and the USSR. Because these estimates differ significantly from previous estimates by this Office, the report also describes the general methodology for the derivation of the estimates and explains the basis for changing them. Section II of the report therefore describes tbe general methodology and rationale underlying the new estimates, and Section III gives Uie new comparisons of GNP and its end uses in the two countriesone In dollars, one in rubles, and an average of these twomod shows bow these comparisons differ froa the ones previously made. In addition, comparisonsore restricted set of expenditures -termed national policy expendituresare shown; they include annual outlays on educution, civilian research, foreign aid. Industrial anil defense. Byreface, some general remarks are In order about the meaning tbat can be attached to International comparisons of GKP's and about the fundamental statistical problem involved inthe GKP's of two countriesocoon currency.
A. Some General Considerations in International Comparisons of Gross National
Some cautionary words axe in order with respect to the meaning and conclusions that should be drawn fron International comparisons of outputs valued at constant prices, such ao those for the Unitedthe USSR presented in this report. First, what is measured ia the relative monetary valuations of goods and services produced and not their relative utilities. Consumers ln the USSR, in contrast to tbe United States, do not exercise any significant degree of sovereignty over the pattern of consumption that is providedand even If tbey did, the quantitative comparison of consumption in the two countries would not, as ailgree, represent relative consumer satisfaction.
This familiar theoretical point holds also for investment and defense. The dollar value of Soviet investmentiven year does not measure the growth-inducing potential (that is, utility) of that investment compared with US investment or even with Soviet investment in other years. Tho critical measure of growth potential is tho additionalyieldednit of new investment, which may be very different in the two countrieo and may vary greatly from year to year In the same country. Finally, statistical comparisons of expenditures on defense (military and space programs) arc particularly elusive in meaning. Although military expenditures may be roughly equal In the United States and the USSR when Soviet military expenditures are valued in dollars, this does not mean that the military establishments In the tvoare equal ln effectiveness or in military potential. Theare made ln terms of price ratios that reflect relative costs of various kinds of military outputnot firepower, mobility, or other attributes of military power. Moreover, some aspects of Military power, such as past expenditures or geographic situations, do not enter into the comparisons at all.
A second caution is that the dollar value of Soviet end-use components docs not measure "cost" to the Soviet economy in the sense, for example, of measuring the share of resources going to investment or the "burden" of defense. The concept of "burden" or share of resources can be measured for each country only in its own domestic costs. For the United States, domestic casts can be approximated by prices. In the USSR, where prices are set arbitrarily by the State, real costs cannot bc estimated, and the burden of defense in the sense of consumption or investment foregone is susceptible of no simple quantitative BKnaure.
3. The Index Number Protlca
The necessary condition for comparing two combinations of heterogeneous outputs Is to express them as value totals using the same unit values or prices for each kind cf output. raditional procedure has been to convert the OS? Ic domestic currency values of one country Into the currency of another country by the International currency exchange rate. rocedure Is meaningless when the exchange rate is arbitrarily set and foreign tradeightly controlled state monopoly as in the USSR. Even in the case of predominantly private trade between two private enterprise economies, the exchange rate reflects prices only of goods and services that are traded internationally and hence is not representative of the full range of output included In GNP.
In this report the outputs of tbe United states measured ln dollars and of the USSR measured in rubies ared In common value terns, using calculated average ruble/dollar ratios that reflect the relative Internal domestic prices Tor identical goods and services in the two countries. The general procedure used to calculate the
average ruble/dollar ratios may be illustratedescription or the derivation or the ruble/dollar ratio lor the rood subcomponent or con-sjwption. First, the pricesepresentative sample of food itemsere collected in both countries, and the ratio of the ruble to the dollar price of each item was calculated. Two averages of the ruble/ dollar price ratios for these U8 items were then calculated, using as weights the expenditures they represent in the United States and in the USSR, respectively. The average ruble/dollar ratio weighted with US expenditures was then used to convert tho dollar value of total US food production to rubles. Similarly, the average ruble/dollar ratio based on Soviet expenditure weights was used to convert all Soviet foodin rubles to dollars. In the former case the average ruble/dollar ratio is said to be US-weighted and in the latter, Soviet-weighted. The two average ruble/dollar ratios are different because the expenditure weights in the two countries are different. For example, consumption of potatoes represents almostercent of Soviet expenditures on food butercent of US expenditures-
In comparing the size of GNP,omponent of GNP, of the USSR with that of the United states, two calculations are thus possible, corresponding to the two average ruble/dollar ratios. Soviet GNP may be converted to dollars by the Soviet-weighted average dollar/ruble price ratio, or alternatively, US GNP may be converted to rubles by the US-weighted average ruble/dollar price ratio. The GNP's of the two countries thus may be compared either in dollar prices or in ruble prices. Since the two average price ratios are different, the two GNP comparisons will be different. This differenceeflectionamous statistical paradox known as the index numberime-honored resolution of this problem is to use the geometric average of the two, equally valid, comparisonsingle number is desired.
The index number problem arises in comparing the GNP's of the -vo countries because their patterns of output and prices differ greatly, reflecting differences in tastes, leveis of income, natural resources, technology, and state of development, and because the relative quantities of goods produced and their relative prices ore different in the two countries. All international comparisons ofncluding the present report, have found that the ratio of prices of goods between two countries ls inversely related to the ratio of the quantities produced. In other words, in each country goods that sell at low prices tend to be produced in large quantities, and vice versa, and these goods are not the same ones in both countries. In the United States, for example, consumption goods are cheap in comparison with Investment und defense goods, and the United States produces more of the former than of the latter; in the USSR, the situationst the opposite.
* For serially numbered source references, see Appendix C.
With reapect to the comparison of GKP's the effect of the inverse relationship between relative prices and relative quantities produced between two countries is to show the size of one country to be larger compered with another when both CKP's are valued ln the other country's prices. This direction of difference arises because, In converting the GSP'sommon currency, the relatively higher prices of the latter country are applied to the relatively larger quantities produced in tbe former, and vice versa. Thus, in the case of the comparisons of GNP and its corcpocenta In the United States and the USSR, relatively high ruble/dollar ratios for consumption are associated with the relatively small size of Soviet consumption, and conversely, relatively low ruble/ dollar ratios go with the higher relative size of Soviet Investment, defense, and administration. esult, Soviet GNP ls shown touch larger percent of USercenthen the comparison is made :ih dollar* than when It is made inercent.
C. The Meaning of the Two Comparisons and of the Geometric Average
-Strictly speaking, neither comparison really measures the difference in total output Of the tvo countries, because no rigorous comparison of the total outputs of two countries lo possible, except in the unlikely cane that they are producing an exactly proportional mix of products. Since in practice the mix of products ls different between countries and can be rendered In caeapexable terms only by valuing the products Of one country in the prices (that is, the scarcity relations) of the other, the ecooamlc meaning of the comparison Is inypothesis about what would result if one country were to shift its production to a. mix of products proportional to the mix of products of the other country. ?/
ilfr'?*! r*.
This economic meaning of an international comparison is based on the presumptionountry can shift its pattern af production without significantly changing its existing relative coots or prices. Thus'/omparison of US and Soviet GNP's measured In dollars, the implicit assumption la' that the-United States could shift to the Soviet pattern of production and still produce the case dollar value of output as before. On thla assumption the cccporison ir. dollar prices is sn approximate measure of the relative ability of the two countries to produce the Soviet mix of output. The comparison of GNP's In ruble priceseasure of their relative ability to produce the US mix of output. The quantitative resultsthat the comparison In dollars is more favorable to the USSR and the comparison In rubles more favorable to the United Statesreflect the fact that each country is better equipped to produce its own pattern of output than that of the other country.
C-P
The geometric mean or average of the dollar and rubleis traditionally usedingle number is desired.* Although the geometric mean may appear toonfused hybrid of two sets of relative prices, its economic meaning is as clear-cut as the meaning of either of the original comparisons. The geometric mean of the two comparisons of the CHF's of tvo countries is an approximate measure of their relative ability to produce an intermediate mix of output, midway between the actual mixes of the two countries. Looked at inway, the geometric meaneasure of the production capabilities of two countries that is less biased than either of the comparisons in national prices.
These Implicit assumptions are, of course, not entirely The dollar comparison assumes that the relative costs ofthe various kinds of output would not change If the United States tried to transfer resources so as to produce the Soviet mix. To makeransfer, however, the United States would have to give up increasingly large amounts of output in exchange for the output more characteristic of the Soviet mix. The more the United States tried to duplicate the Soviet production mix, the more costs of production, in all probability, would tend to rise. esult, if the United States produced the same product mix as the USSB, it would not show up as well relative to the USSR as it actually does in the dollar comparison madeCPO Similarly, the ruble comparison assumes implicitly that the USSR could shift to the production of the US output mix with no cnanges in unit costs or prices. Such an outcome would be most unlikely.
hu USSR attemPted to reduce the same output mix as the United States, the USSR would not appear in asight as it does in the ruble comparison shown in this report.
The dollar comparison and the ruble comparison thus overstate the relative ability of the US and the USSR, respectively, to produce the Soviet and the US mixes of output, respectively. As already stated, the traditional resolution of this problem is toeometric average of Uie two comparisons. It should be noted, however, that in comparisons of two countries at greatly different states of development, sucn as tne United States and the USSR, the geometric mean tends tTovcrstate tic relative position of the less advanced country. The reason Is that'the more advanced country undoubtedly could shiftess coool^cated product mix with less additional cost than would'accrue to the lesscountry if it were to shiftore complicated mix.
The geometric average Is used in preference to the arithmetic because eneral, economic grow Li: and other changes proceedany: that is, constant percentage increases describe the Changes better than constant absolute increases.
i.
II. Method of Co.') struct ing the Comparisons
This reportomparison of GNP and its major end uses in the United States and the USSR for the. GNP consists of the aggregate value of all final sales, including the expenditures of households on consumption; the expenditures of government for health, education, administration, and defense; and the expenditures ofgovernment, and producing enterprises for gross investment. Toistribution of GRP by end use, these expenditures are grouped into four major categoriesonsumption, investment, defense, and administration. Tablesresent the estimates of US and Soviet GNP and its end uses in rubles and in dollars for each of the The comparisons are made5 rubles, because information on Soviet prices is not available in the necessary detailore recent year, and3 dollars,3 is the most recent year for which the US data are available in the required detail. The general method used was first to calculate Soviet GNPn5 rubles and US GNPn3 dollars, with the total GNP for each country broken down into comparable end-use categories; and second to convert US GNP and its end uses5 rubles and Soviet GNP and its end uses3 dollars by means of5ollar price ratios. These average ratios were calculated separately for defense, for administration, and for four subcomponents of consumption (food, nonfood goods, consumer services, and health and education) and three subcomponents of investment (machinery andconstruction, and inventories). The selection of these particular categories was determined primarily by the availability of US price indexes for derivation of5ollar ratios and for conversion of US GNP from current dollars3 dollars.
A. Estimation of Gross National Product and End Uses
Unlike the United States and most other industrialized countries, the USSR does not publish estimates of GNP. Published Soviot data are couched in terms of the Marxian concept of "material,hich differs substantially from the-Western definition of GNP. Soviet GNP, therefore, had to be estimated independently from the inadequate published official statistics andariety of other bits and pieces of The procedure required the laborious calculation of ruble values for dozens of individual components and their aggregation into major end-use categories tnat conform as nearly as possible to Western definitions. The basic estimates were calculated for the55 ruble prices.** For the, the values for the end-use categoriesonsumption, investment, defense, and administration5 prices were obtained by use of end-use volume indexes constructed
" "i;
** Tor detailed sources and methods used to estimate Soviet GNP, seend the footnotes to Table 8.
for the individual components. Estimates of total Soviet GNP5 ruble prices for theowever, were obtained by moving5 estimate forward by the weighted sum of indexes of GNP by sector of originndustry, construction, agriculture, transportation, communications, and services. Such an indexore accurate volume indicator of year-to-year change in total GNP than does an index basedummation of end-use components because of the inability to measure accurately (l) inventory change in real termset foreign Investment. Although the two indexes do not differ significantly for the periodhole, as can be seen in Tablehe sector-of-origin approach gives values for total GNP in some years that differew percentage points from those obtained by summing the four end-use com-
For the United States, estimates of total GNP and itscurrent dollars are published by the Department of Commerce.* in current dollars were expressed3 dollars by meansprice indexes for subcomponents of the end-use categoriesbe derived from data given in publications of the Departmentand the Council of Economic Advisers. The breakdown of GNPend uses as shown in Tableowever, differs from thatthe Department of Commerce, because the data for variousGNP had to be reclassified among the end uses in order to obtaincategories comparable in definition to those permitted bydata. The alternative procedure of adjusting the Soviet datathe US definitions could not be employed, because of theof published Soviet statistics. The reclassification of thecategories of US GNP is presented in Tablend the derivationsubcomponents of consumption and investment is explained in Tabledefinitions of each end use as referred to in this report are
Consumption
This component includes total household expenditures on goods and services plus government noninvestment expenditures on health and education. This definition is required i'c- comparability because in the USSR nearly all outlays on health and education are made by thewhile in the Unitedignificant share of expenditures for these purposes is made by households.
* The GNP data for the United States are those published prior to the revision in national accounts made by the Department of CommerceInformation on the revisions has not as yet been published in sufficient detail to permit their use in these comparisons. In any case, the nature
is. riat V: c ould
significantly affect the US-USSR comparisons of GNP and its major end uses
** Pp.espectively, below.
Investment
This component includes gross investment in construction and in machinery and equipment as well as changes in inventories. It includes government stockpiling and investment In defense production facilities but not direct military construction and military equipment purchases, which are included in defense. Also included in investment for the United States are expenditures of state and local governments for the development of atomic energy, net foreign Investment, and military assistance; the latter two items are not included in the figures for the USSK, however, because of lack of the requisite data to estimate them. To obtain comparability with the United States, expenditures on capital repair and civilian research and development. Included under investment in published Soviet statistics, were deducted from Soviet investment. In the United States these expenditures are included in current costs (and prices) of other outputs and not aa separate final output.
Defense
The defenae category includea (l) pay, subsistence, and other current operational expenditures of the armed forces (includingsecurityilitary construction and equipmentilitary research and development (including all space) expenditures; and (U) atomic energy expenditures. Military pensions were excluded for both countries because they are merely transfer payments.
Government
For both countries this category isesidual of current government outlays on goods and services not included in the three categories listed above. The figures for the USSR exclude the cost of administering state-owned enterprises because these overhead costs of enterprise management are included in product prices and appeur, as in the US accounts, in the value of the final uses to which these products correspond. Tbe estimates for the USSR, however, do include expenditures by the Communist Party, which servesey arm of government administration and control; also included arefor nonnllltaxized internal security activities, some of which have no counterparthe United States.
B. Derivation of Ruble/Dollar Ratios
Having obtained estimates of GNP by end use for the twothat are comparable in definition one for the united States3 dollars and the other for tha USSR$ rublesit is then necessary to express the estimatesommon currency in order to measure the relative positions of the two countries. The next step in the derivation of the comparison, therefore, was to obtain two sets
of appropriately weighted, average ruble/dollar price ratios, one set with US weights for converting Soviet GNP by end use into dollars and the other set with Soviet weights for converting US GNP by end use into rubles.* The desired ratios were derived in three steps: (l) basic ratios5 were calculated as accurately as possible with available data for the four subcomponents of consumption and the threeof investment previously listed, for defense, and5 price ratios were converted5ollar ratios by dividing them by appropriate US priceor reasons that will be explained below, certain of these basic ratios were raised arbitrarily to correct in part for known biases in the data. The derivation of the final ratios is shown in*
1- Calculation of the Basic Ratios
For consumption and for investment, the prices of asample of presumably identical goods and services in the two countries were collected for the5 (when the most information for the USSR isuble/dollar price ratio was computed for each product or service by division; the ratios for the individual products and services were then grouped into subcomponentsfour for consumption and three for investmentnd two average ratios for each subcomponents were computed by weighting the individual ratios by the value of sales5 in each country; and these average ruble/dollar ratios5 were then converted5ollar ratios by dividing them by appropriate US price indexes. For defense, the ruble/ dollar ratios were derivedyproduct of the calculation of defense expendituresomponent of GNP in the USSR. These expenditures were estimated5 by valuing the various components of Soviet military programs both5 rubles and3 dollars. For the, dollar and ruble valuations were obtained by moving5 estimates forward by physical volume indexes. Ruble/dollar ratios for defensehole and for individual components in each year were derived from these valuations by division. US defense expenditures in dollars were similarly estimated for individual components and converted to rubles using the ruble/dollar ratios obtained in the valuation of Soviet programs. For administration, the ruble/dollar ratio is basedingle comparisonnamely, the average annual wsges of employees ii: government administration in the two countriesonverted5 rubles3 dollars by dividing the ratio5 by an index cf average wages of government employees in the United States.
2. Adjustment of Certain Ratios
In deriving the final ratios used to convert the endS and Soviet GNP to common currencies, several of the basic ratios
"ore detailed explanation of the derivation of the basic ruble/ dollar ratios, secnd the footnoUd to Table, beiow. elow.
escribed above were raised arbitrarily. The adjustment of these ratios represents an attempt to allow, albeit in an arbitrary manner,ias ln the data that cannot be measured precisely but that tend* to overstate the position of the USSR relative to the United States. In other woras, it is an allowancenown biasnown direction, but of unknown size, in the comparisons asf relative production capabilities of the two countrlea. The adjustment is preliminary and necessarily imprecise, but is regarded as minimal in the senseorrect adjustment arguably could be much larger but is not very likely to be smaller. The consiaerations underlying theare both general, relating to the nature and meaning of the basic data themselves, and specific, relating to the rationale for adjusting the Individual ratios.
a. Uncertainties and Biases in the Data
Rach step of the calculations described in the sections above involves Inadequate data. In assessing these calculations and the nature of the Soviet data involved, the most illuminating hypothesis to be borne in mind ls "Murphy's law"hatever con go wrong will go wrong. Because of the nature and shortcomings of national Income data published by the USSR, an estimate of Soviet GNP must be pieced together from announced budget and expenditure data that are incomplete,and only dimly understood, for example, in the Soviet national budgetillion rublesercent of the total) were simply unidentified as to purpose. This amount had to be allocated somehow or other between final product expenditures and transfers or disguised subsidies to producing enterprises. Furthermore, far* income Inarge part of consumption, mist be quite independently estimated, and the data with which to do this are extremely meager.
The sktapiness of Soviet announced data, however, is perhaps lessroblem than is interpretation of the nature of the economic activities that the data purport to measure. This report seeks to compare the values of the final products of the two economies. However, what are final products in the USSR is far from obvious. This questionn the accounting sense what sales aro sales to final usersn the physical sensewhat are the exact goods and services which these sales consist of. An iridispeniiablc auxiliary question lst what prices are these goods and services actually entered into the reported sales. entrally planned economic system shapes tha onswero to these questions quite differentlyarket system.
The scarcity and the ambiguous character of Sovietdata for the most partargin of error that aay fall in either direction. In other words, the estimate of Soviet GUP in rublesange of uncertainty,vever, ls probably small. The estimates in this report agree fairly closely with independentmade by several Western students of the Soviet economy.
- is -
Although the estimate of Soviet GNP in rubles is considered to be reasonably satisfactory, the same cannot be said for the basic ruble/ dollar ratios. Every attempt was made to match products in the two countries as exactly as possible, but the data problems encountered in the attempt are such as to produce ruble/dollar ratios that clearly overstate the value of Soviet output compared with US output. These problems relate primarily to the physical composition of the outputs (individual products) being compared. Following ore illustrations ofew of these problems. In calculating the ruble/dollar ratios for individual durable goods, it was necessary in many instances to match Soviet products with US equivalents that are obsolete: in other instances, Soviet products had to be matched with superior US models Moreover, the quality of workmanship and the product durability of Soviet products were assumed to be those of the listed specifications, whereas actual performance is known to be greatly inferior to norms and standards. Furthermore, many types of equipment sales in the United States are accompanied at no extra chargearranty, initial troubleshooting services, training of operators if need be, and help in planning the most efficient integration of the equipment into theprocess.* These auxiliary services are usually missing or ignored ia the USSR, and their absence makes otherwise identical physical products different, less costly, and less useful. To some degree all theseimply that the Soviet goods are inferior to the US counterparts and tnerefore that their equivalent dollar cost is less than shown by their ruble/dollar ratios. These problems are present also in the comparison of construction in the two countries.
overall mix and pattern of Soviet production also
leadsample bias in favor of the ^SSR. Soviet planners haveon the mass productionelatively small list of simple general-purpose goodsthree or four models of cars, standard machine tools, small uniform apartments rather than single-family urban houses and so on. The absence of diversity of models andariety of specialty equipment is Characteristic in the Soviet mix of producers' uuraoles.** Ruble/dollar ratios based only on such standard items clearly are inadequate indicators of the relative ability of the two
-
Kith respect to producers' equipment, however, it should beince the prices compared In this. producer's "tion, net of packing, auxiliary services are not explicitly in the price, although some may be included implicitly. Their exclusion would be another example of sample bias discussed below. The same can oe said of transportation and distribution charges, which are excluded irom the price comparison, and which evidentlyarger percent of purchasers' prices in the US than in the USSR. This raises thethat these are more useful but costly services which are being neglected or inadequately performed ir the USSR.
The sample of products used to derive ruble/dollar ratios thuslimited almost exclusively to standardized,and does not take account of the variety of specialty-.ooms large in the US product. ir.ix and thai would be verv costlvin the
-
industries to produce. ThiE bias Is reinforced by the nature of the available sample of Soviet prices, which consists almost entirely of shelf items; prices for such special-purpose items as were produced are unavailable.
One final problem occurs in the use of costs (wages and materials) to measure tbe value of output of services in education, health, and goverrjaent administration. In the OSSB these costs consist largely of wages and salaries. The basing of ruble/dollar ratiosomparison of average wages, as is done in this report, assumes equal productivity of labor in the two countries and thus clearly overstates the output of services in the USSR relative to the United States,
b. Justification for Raising Individual Ratios
The statistical problems stemming from the deficiencies of data described above axe insoluble by the analyst making these The data upon which to base precise adjustments for quality, longevity, diversity, and the mix of output are not available. Those aspects of incomparability of products that could reasonably be allowed for have already been taken into account in calculating the basic ruble/ dollar ratios. The problem that remains is to allowias of known direction (ln favor of the USSR) but of unknown size. The solution adopted here is to raise the ruble/dollar ratios most affected by the biases by on arbitrary, uniformercent. The ratiosn consumption, the ratios for radio and television equipment, electrical and other appliances, and automobiles in the subcomponent nonfood goods, and the wage component of the subcomponent health andn Investment, the ratios for machinery and equipment and for construction;he ratio for government administration. omparison of original and adjusted ratios, sec)
With respect lo consumer durables, the difficulties of matching Soviet and US products were such as toonsistent downward bias to the ruble/dollar ratios.** In general, Soviet durables such an automobiles, washing machines, and refrigerators resemble models produced in the United States before World War II. SoviEt products had to be compared, for the most part, with the cheapest US models, and even tliese were generally superior to the Soviet goods in appearance,and operating features of one kind or another. Left out of the comparison entirely, for lack of Soviet counterparts, were thevarieties of modern durables available to US consumer*uch as fully automatic washing machines, clock-radios, dryers, and the likewhich would be relatively high-cost products in the USSR. Although comparative measures of longevity and durability arc not available, it
P.elow.
** These problems arc described product-by-producteport by this Office that compares consumption in the United States and the USSR (see.
-
seVv^lST TnLf^ of complaints in the Soviet press aboutcon*tant breakdowns, and unsatisfactory repair services
Sv ofoviet counterparts Byra"on' thc ^ess reports thatecent0
and ^Jm0delMoscow customers,
^OT^PairStbC8eWerein a L Press also provides much evidence that
fd te^vision sets are inferior to US products in reliability and general performance, j/
,,reaPect to machinery and equipment the evidence
available from the Soviet press and from the reports of foreign observers and users of Soviet products shows clearly that theirS inferior to that ofurables in the United States {hi"evi
mTin^'V?CrieSrep0rtB bvealing vith individual industries. Soviet jet aircraft engineshort
*th^ the Unitedodels of
and eraln combines arehe service UfTo?
SPUnP! USed in thC cheoical indu=try is measured in hours
"eryCC^"SS0"be recced
tnTrL^.. Soviet-produced tires average from one-half to two-
3 eff^ient, less durable, and more
S^mX ?Pfate than WeStern equipment. ecent US delegationiet fflachine Cools conl-inue to be less durabL
ess *recise than
Sovfet^ fCe developed countries that haveIfaent Provides additional evidence. Thus the teeth of shovel*
In Iraat0ou" of excavation; traces
cSL7 uT overhaulours of operation compared with
i'noVJJZ aS. trtT" 'T'T1eclmc
iuTictil problems were chronic
with jeeps trucks, generators, and loading cranes; in Cuba, blades of
ut of'lO buLdozer allegedly
ctX brokedownnd iRProvinceu' 6 than 2and in EgyPt and Afghan-
efl^encieeurability and operation^ere experienced with respectide range of heavy construction
ences 0 lnadeo-uate allowances for quality'ratios are biased downward because they do not Itl and Zr^roduce efficiently the com-Swduct ,ff hnt that are very important In the US product mix but are not represented at all in the sample ofotios. The possible magnitude of thiS bias may beT^aSa^
^ IV thCKhru-hchev's This study compares the ruble cost cf building the planned
-
culuted dollar cost of the same capacities anduble/dollax
ratioomparedatioor all industrial investment baaed on the sample of products used In this report.
The low quality of Soviet construction is well known both to the Soviet citizens tbemsolves and to Western observers. The evidence ls summarizedcries of reports by this Office. Ik/ Tbe Soviet press Itself in words and photographs provides the best sourceudgment as to tbe coaparative quality of US and Soviet Tbe problems range from poor materialslarge percentages of broken bricks, poorercent below US qualitynd excessive breakage of window glassto poor construction workbad finish, plumbing, and wiring and frequent structural failures and collapses. elegation of construction experts that visited the USSR recently concluded that Soviet construction was "unbelievably" bad; one member of the delegation stated that with few exceptions the construction projects that they saw would not pass inspection in the US. So shoddy vas tbe work ln many buildingsreat deal of expensive repair and replacement would be needed to keep them habitable.
In the case of administration and of the wage component of health and education services the adjustment of the ruble/dollar ratios represents an allowance for differences in the productivity of labor ln the two countries. For both categories, outputs are closured in terms of inputs, largely labor, and the ruble/dollar ratios arc mainly wage ratios. This procedure assumes equal productivitythat is, equal quality of these inputBclearly an incorrect supposition. The physical capital equipment available to US workers in health, education, andexceeds that available to their Soviet counterparts probablyarge margin. In addition, the average educational levels of th* two groups are quite different. The average US worker employed ln health services had5 years of schooling, comparedears for the average Soviet worker in that field. For workers in the field of education the corresponding achievements2nd in government administration they5. Although directof relative productivities cannot be made for these services, come notion of the probable magnitude of the difference is given by the fact that in industry, output per worker in the USSR is only about ono-third of that in the United States and that ir. the economyhole, as neasured by GNP per person employed, productivity is onlyf the US level. Finally, with respect to administration it should be noted that the ruble/dollar ratio does notaterialsbecause of lack of the necessary data. Inclusion of materials would raise the average ruble/dollar ratio considerably, as vas the case in the calculation of the average ratio for health and education.
3. Pinal Ruble/Dollar Ratios for Gross National Product nndJaes
Th*ollar ratios for theof consumption and investment, lor defense, and forwere then used to convert each of these components of US and Soviet
DBF to rubles and to dollars, respectively, for each yearotal values for consumption, for investment, and for GNPhole (United States in rubles and USSR in dollars) in each year were obtained by addition of the appropriate components. Finally, ruble/dollar ratios for consumption, investment, and total GNP were calculated from tneruble and dollar values ror tne two countries.
The US-weighted and Soviet-weighted ratios, together with their geometric averages, for GKP and for each end use arc shown for illustrative purposes in* These average ruble/doilsx/ratios are tha result of weighting appropriate subcomponent ratios separately for each year ln the comparison by their respective shares ln GNP in each year. An alternative procedure would have been to carry out theusing base-year The use of given-year weights, ln fact, makes little difference. The geometric average of the two ratios for consumption declinesC5 weights0eights. The geometric average of the ratios foralsoittle5. Tlie ratio for defense is the most sensitive to the weights used, with the geometric average ranging35 weights94 weights. The ratio for administration remains constant The overall ratio for total GNP varies by only two one-hundredths as the weights change over time.
In any given year, however, the Soviet-weighted and US-weighted ruble/dollar ratios for GNP end far most of th* and uses differ greatly. The extent of this difference between the two ratios varies considerably among the components, being greatest for consumption and least for defense. Among the subcomponents of consumption andthe largest differences in tbe Soviet and US-weighted ratios are ror health andomparedad for machinery andompared. In tbe case of health and education the large difrerence results from greatly differentor factor inputs ln tbe two countries: in these fields the USSR, compared with the United States, uses more labor, which islow poid in the USSR, and less materials, which ore relatively high priced there. With respect to machinery and equipment the large difrerence ln the two ruble/dollar ratios results from the great divergence in relative prices and pattern of output for producers' durable equipment ln the United States and the USSR and the inverse correlation between the output ratios of the two countries and the corresponding price ratios. For example, the United Statesthe USSR by Tar in automobiles, the ruble prices for which are high relative to US prices. On the other hand, the production of machine tools is relatively more important in the USSH than in the United States, but the ruble prices for machine tools are lower relative to dollar prices.
elow.
II. Comparisons of Relative Size of US and Soviet. Gross National Product aad Major End Uses,tfe
A. Coaparative Size of Total Gross Hational Product
Tbe final percentage carparisons of GKP in tbe United States and tbe USSR for thepresented Innd Tablehe table gives three percentage comparisons | one Is obtained vhen thef both countries are valued in rubles, another is obtained when the GNP's ore valued in dollars, and the third is the geometric overage of the ruble and the dollar comparisons. The ruble and dollar values are shown ln Tables* According to the data in Tableoviet GNP5 wU less than one-third the US level when measured in rublesittle less than one-half the US level vhen measured In dollars; the geometric average shows Soviet GNP to be somewhat less than two-fifths of US GNP. Soviet GNP increased rapidly ln sizeto that of the United States, lost ground9 increasedigh ofercent (geometric average) of US GNPnd then declined slightly ln relative size. Thus Soviet GNP has not guined on US CNP oviet GNP wasercent of tbe US level by the ruble comparisonf the US level by the dollar comparison; the geometric average isercent. er- capita basis, Soviet GNPmaller percent of US GKP than when total GNP's are compared. InSoviet per capita GNP was slightly more than two-fifths the size of US per capita GNP in dollar prices and about one-fourth an large in ruble prices. Soviet per capita GNP Increased in size relative to the US over the, except for slight declines9 oviet per capita GNP was nearlyercent of the size of U9 per capita GNP at dollar prices and aboutercent as large In ruble prices. In the geometric average comparison, Soviet GNPer capita basis wasercent of US per capita GNP8 percent of US per capita GNP
From an analysis of the average annual rates of growth of US and Soviet GNP, shown ln Table 2, it is clear that theincrease ln the relative size of Soviet. GNP in thes largely attributableonsiderable decline in the US growth rate rather than to any substantial acceleration in the Soviet rate, which was quite high during the three years. The decrease in coaparative size of Soviet GNPas due toubstantial slowdown ln the rate of growth of Soviet GNPonsiderableof the US growth rate During the five, Soviet GNP Increased slightlyercent of US GNP, because, although the Soviet growth rate continued to decline, the rate of growth of US GNP declined somewhat more.
* espectively, beiow.
-
Soviet Cross National Productercentage of US Gross National Product
USSRercent of US
Ircss national product
Ruble coaparlaon
Dollar coaparlaon b/
Geometric average comparison cj
Consumption
9
9
O 5 2
6 76 9 0 1 2 k
0
7 2 1
comparison 0/
Dollar eoaparlson fi/
Geometric average coaparlaon cj
6 1 6
0
5
6
8 5 2
coaparlaonollar coaparlson b/ Ceocctrlc average covajarlion c,
f*-
7 2 4
66.
3
2
1
781
comparison a/
Dollar comparison b/
Geometric average coaparlecn cj
Administration d/
Dollar comparison b/
2
C
0
7
2
1
3
7
9 27 0 62 1 3
values of Soviet expenditures in Tableelov, divided by the ruble values of US expenditures laeiow, and the quotient expressedercent.
values of Soviet expenditures Icelov, divided by the dollar valuesexpenditures In Tableelov, and the quotient expressedercent.
geoaetrlc mean of tha ruble and dollar comparisons.
a. Only the dollar comparison is presented; the ruble comparison vould be the some oecauseand UQ-vatfthted ruble/dollar ratios used for administration areinir 11 1
i s js n 33 s ss
3 33
a5
ij 3
f! j
ij f
>'
35 35 33 35 33
I, , . jjj
jiijMliMiill
US and USSR: Major End Uses of Gross National Product
IA
4
B. Comparative Size of Major End Uses of Gross National Product
The relative sizeB of the US and Soviet economies differin regard to the major end-use components of GNP, as shown by the dollar, ruble, and geometric average comparisons of consumption, investment, defense, and government administration presented lnsee also The relative positions of the several end uses ln the USSRis the United States also have changed markedly over the comparatively short. Thus, whereas5 Soviet consumption and investment both were in the neighborhood of one-third that of US levels,4 investment had risen to more than two-thirds of that of the United States, but consumption was still not far from one-third that of the United States. In contrast, the relative positions of defense and administration declined, the formermore rapidly than the latter. On the whole, these trends in the relative positions of the four end uses have proceeded fairly smoothly over the period, although there have been sharp changes In some years inone or another of the components.
Soviet consumption was about one-fourth of US consumption5 at Soviet prices and about two-fifths at US prices. During the next eight years, Soviet consumption increased only slightly faster that US consumption and4 wasercent of US consumption at Soviet prices andercent at dollar prices. In the geometric average comparison, Soviet consumption wasercent of5 andercenthe corresponding percentageser capita "basis arc The largest increase in the relative size of Soviet and US consumption occurred in the. 9 the relative size of Soviet consumption declined slightly from the previous year, then increased again moderatelyndslightly3 The annual rates of growth of US consumption expenditures are quite steady for the, so that the slowdown in the gain in relative size of Soviet consumption during the latter years of this period Is due almost entirelyecrease in the rate of growth of Soviet consumption.
Although Soviet consumptionhole approached two-fifths of the US levelotal expenditures of private households on goods and services were lesshird of those of US consumers; expenditures on health and education in the USSR (largely by theowever, approached two-thirds of outlays on these services in the United States. Private household consumption in the USSR gained slowly relative to the United States5ut expenditures on health and education relutive to those ln the United States declined slowly59 and then rose again to occupy4 the same relative level as5
The size of Soviet total investment relative to that of the United States changed markedly. oviet investment washird the size of US investment at ruble
prices, two-fifths os large at dollar prices, and nearlyercent as large according to the geometric average. In the, Soviet investmentercent of US investment almost doubled, and8 Soviet investment was nearly three-fifths of US investment at ruble prices and three-fourths of US investment at dollar prices, theaverage being two-thirds. The relative size of Sovietdeclined9 and then rose toercentnd rose again3U, when its relative position wasittle higher than The spectacular increase in the relative size of Soviet investmentesultedery high rate of growth of Soviet investment(nearlyercent annually) and from an actual decrease in US investment. The much smaller increase in the relative size of Soviet investmentas dueubstantial decline in the rate of growth of Soviet investment expenditures (approximately 7annually) and to an increase in the growth rate of US investment. The spread between the ruble and dollar comparisons for investment is smaller than that for consumption, indicating less difference in the Soviet and US price structures and output mixes for investment goods than for consumption goods.
Comparisons may also be made with respect to fixed investment, which makes upoercent of total investment in botn countries. It represents the outlay on new plant and equipment and is calculated as the sum of construction and machinery and equipment. In most years, Soviet fixed investment is somewhat larger relative to US fixedthan is Soviet total investment relative to US total investment, but the general trends and relationships are about the same for the two measures of investment.
The estimated size of Soviet defense expenditures relative to US defense expenditures declined significantly. As measured by the geometric average of the ruble and dollar comparisons, Soviet expendituresercent greater than US expendituresuthe Soviet defense effort wasittle over four-fifths the size of that of the United States. The relative size of Soviet defense expenditures decreased sharply6esult of their absolute decline in these years, and7 Soviet expenditures had fallen to less than nine-tenths of US , Soviet defense expenditures rose slightly relutivc to those of the United States, but declined thereafter. In this period, Soviet defense expenditures rose, but US expenditures rose faster. Over the eight-year period the difference between the percentage comparisons in rubles and in dollars narrowed substantially, reflecting the fact that the Soviet and the US mixes of defense expenditures have become much more alike. The USSR has shifted more and moreanpower-intensive defense effort to on increasingly sophisticated and relatively expensive raoderr: weapons program. Soviet defense expenditures relative to those of the United States
-
decreased considerably more slowly according to the ruble comparison than according to the dollar comparison. This difference indicates that defense costs in rubles are rising at on increasing rateis the United States, as the USSR shifts more and more toward an output mix in which Its competitive advantage is less.
Inasmuch as the US-weighted and Soviet-weighted ruble/dollar ratios for government administration arc the same, comparisons ln rubles and in dollars are the same. Over the(hole, Soviet expenditures declined relative to US expenditures, but there were marked fluctuations within these years. 5 Sovietfor administration were three-fourths of US expenditures. The proportion rose to four-fifthseclined to about two-thirdsose to nearly three-fourths inand declined again to two-thirdsI*.
C. Comparative Size of National Policy Expenditures
ndndresent estimates of the comparative size of US and Soviet outlaysn several strategic components of GNP-which may be grouped under the rubric of national policy expenditures. They include education, industrial investment, civilian research and development, foreign aid, and defense. omparison assesses the relative position and progress of the USSRis the United States in areas that directly reflect the declared policies of the USSRhe achievement of rapid economic growth, especially in industry, snd the promotion of foreign policy objectives through foreign aid and an expanding defense establishment.
Comparative expenditures on the first threeducation, industrial investment, and civilian research andreflect, in general, the relative commitment of the two countries to the maintenance of end increase in present rates of economic growth. Comparative outlays on education representinvestments in raising the average level of education of the labor force, an important factor in the rise of labor productivity that contributes to economic growth in modern industrial societies. Similarly, comparative expenditures on that part of investment allocated to the expansion of industrial plant and equipment show the willingness Of the two countries to allocate present resources to achieve future growth; future power positions of the two countries will, depend in part on such present resource allocations. omparison of outlays on civilian research and development is useful for assessing the potential for growth that will ensue from technological improvements and new inventions.
The detailed data needed to calculate national policy expendituresJ1 arc not uvailable.
-
-Mil'- 3
Percentage cf US National Policy
f US
xpenditures
earpariisr. a/
- *
iwipwlsan h/ Collir caiparlsott a/ Geoictrle average coapvlsca c/
3
6
a
lU.lt
ir.feitneiil
&/
Dollarveruee corpurlrwi cf
Jiv! .. 1 ui.!
toiler :cr.par:son e/
a
A
aid rf
.7
conpwison bf
Dollar ccnpirlior b/
v?rese csnpariion c/
7
8
8
7
2
ruble/
The dollar volutin of So*lot expenditures divided byiar values of b. Th* uble values or Soviet expenditures divided br th* ruble values of US expenditures fron The geometric neon of the ruble sad dollar ccnparlecoi.
a. Or-ly Me dollar cowlson la rubla conparliOD would be tha nana tacauta tba Soylat'VBlgntad dollar ratloi used for civilian raaaaich oral development are Identical, e. Available data do not jumlt tbe nattlnguble comparison.
Alongarge and growing investment* in physical capital relative to the United States, the USSR has been devoting impressive amounts of resources to the development of its human capital. NP less than half that of the United States, the USSR spent nearly nine-tenths as much as the United States on education, including both current expenditures on wages and materials and the cost ofof new educational facilities. As shown by the geometric average comparison, the Soviet relative position rose steadily over the period, fromercent of the us5 toercent By the dollar comparison, Soviet expenditures on education exceeded those of the United States in both years. Expenditures on educationuch larger share of total consumption in the USSR than in the United States. 3 they representedercent of the total in the USSR when measured in dollars, comparedercent in the United States; the corresponding percentageshen measured in ruoies. With this investment effort the USSR raised the median number of years of schooling of its population by nearly one-half, fromears5 toears The average educationalof the US population, however, was5 years
The relative size of investment allocations by the USSR to the industrial sector of its economy increased spectacularlyit more rapidly than for investmenthole. According to the geometric average comparison, Soviet industrial investmentf that of the United Statesut8 it almost equaled that of the United States, and3 it wasercent greater. Soviet industrial investment has exceeded US industrial investmenthether measured in rubles or in dollars. The emphasis that the USSR gives to the development of Its industrial sector is shown even more strikingly by the fact that industrial investment in the USSR is twice ashare of total Investment as in the United States. Thus when measured in dollars the USSR3 allocated nearly kOof total investment to industry, whereas the United Statesittle overercent of the totul to that sector.
. Soviet expenditures for civilian research andwere less than one-third those of the United States, but3 theyittle over one-half of US expenditures. The relative size of Soviet civilian research decreased slightly67 but rose again in the. The US-weighted and Soviet-weighted ruble/dollar ratios for civilian research are equal, so that the comparative size of US and Soviet expenditures is the same, whether valued at dollar prices or ruble prices.
A discussion of the relative siee of US and soviet defense expenditures appears in the preceding section. With respect to foreign aid, the USSR has improved its positionis the United Statesreater extent than for any other of the national policy expenditures, but the relative size of this component is also by far the smallest.oviet foreign aid was less than one-tenth as large as US
-
foreign aid, but3 it had increased to almost one-fourth tbe size of US aid. Soviet foreign aid expenditures increased greatly relative to US expenditureseclined8harply during the next three yearseak ofercentnd then decreased toercent The comparison is made in dollars, because the data necessary touble/dollar ratio for foreign aid are not available.
A whole host of Qualifications attaches to the comparisons of expenditures on foreign aid in the two countries. In the first place, the estimates could be made only in current dollars, because of the lack of an appropriate price index to convert them3 dollars Ixjllar figures for Soviet foreign aid had to be pieced togetherariety of sources of varying reliability and do not include military aid to Communist countries. umber of "price" biasesthe estimates weaken the reliability of the comparisons. Because Soviet list prices for military aid items are much lower than prices for comparable US equipment, particularly for aircraft andimple dollar-for-dollar comparison of US and Soviet military aid tends to understate the value of the latter. On the other hand, Soviet technical assistance, which represents aboutercent of Sovietaid, is less efficient and relatively more costly than comparable US assistance, so that an upward and offsetting bias is introduced into the comparisons. In addition, the major part of foreign aidconsists of machinery and equipment, often In the form of complete Plants, for which appropriate estimates of world prices are not Finally some upward bias may exist on the US side because most us foreign aid is given with the provision that the recipient must spend the funds in the United States; it is possible thot US prices for some goods are higher than the world prices on which Soviet aid expenditures are said to be based.
cent.
Takenhole, the five categories of national policy expendituresuch .Larger part of total GKP in the USSK than in the United States. In dollar valuations, total national policy expenditures5 were kO percent of total GNP ln the USSR and onlyercent in the United States;3 their share in Soviet GNP hud dropped toercent, while their share in US GNP had not cnanged significantly. Whereas total GNP when valued in dollarsercent of US GNPotal national policy expenditures in the USSR wereercent of those of the United States. oviet GNP had risen toercent of the US level In dollars, whili
onal policy expenditures had risenesser extent,er-
D* Comparison of Present and Previous Estimates of usGross National
The comparison of US and Soviet GNP and Its end uses presented ir, tms report differs significantly froir. previous estimates by this
-
T J
Office. The new comparison shows the size of the Soviet economy to be smaller relative to that of the United States as measured by total GNP and each major end use except defense. The difference Is attributable to tvo changes introduced into the nev comparisons:, (l) the arbitrary adjustment of certain ruble/dollar ratios for reasons explained above,he omission of expenditures for capital repair and civilian research and development previously counted in the estimates of Soviet investment; those items make up aboutercent of total investment. Soviet expenditures for capital repair are excluded because in large part they consist of outlays for current maintenance and repair. On the other hand, it is believed that sizable amounts of capital repair are not included ln investment reporting in the United States but instead are written off as current costs. Similarly, expenditures for civilian research and development are counted as costs and not as investment in US statistics. The adjustment of the ruble/dollar ratios, however,uch greater effect on the US-Soviet comparisons than does the change in the estimate of Soviet investment.
hows the present and previous percentage comparisons of US and Soviet ONP and its end useshe corresponding ruble and dollar values are shown in* The revisions in theaffect only the levels and not the trends, because in theivilian research and development plus capital repair did not changehare of total Soviet investment and because the adjustment of the ruble/dollar ratios was applied uniformly to all years for lackore defensible alternative. With respect to total GKP the new-comparison lowers tbe relative position of the USSRis the United States by one-twelfthroaercent to kh percent3 by the geometric average comparison. The relative position of the USSR for consumptionhole is reduced by onlypoints, but for health and education lt is significantlyercent compared withercent). As would be expected, the new comparison substantially changes the relative position of the USSR with respect to investment. Measured by the geometric average, the new comparison shows Soviet investment tc beercent of that of the United States, compared withercent according to the previous comparison by this Office. For administration the comparison is changed proportionately to the change in the ruble/dollar ratio; Soviet administration Is shown to beercent of that of the United States instead ofercent.
* elow. The data toprevious" comparison4 are not available in the necessary detail, lt is clear, however, that the differences between the new and old comparisons would bc about the same'! and are shown innd
-l Li '1
iaoie h
f Present ind Previous Estimates of Soviet Gross national Product asPercentage of US Gross National Product3
ercent cfComparlsona
Geometrio Geometric
Comparison Comparison Average Comparison Comparison Average
Gross national product
V
consuaptloa
Nonfood goods Consumer services
9
2
5
9
2
5
and education
Investment
and Construction
0
0
Calculated froa data in Tablesp.espectively, below. Ruble comparisons are based5 rubles and dollar comparisons3 dollars.
APPEKDIX A
DERIVATION OF BASIC BUBLK/DOLIAK RATIOS
A. Conn uic.pt ion*
5ollar ralios for consumption were derivedomparison of US and Soviet pricesampletems, including u8 food items,onfood Items,ervices (excluding health andealth and education services. Although the sample is quite representative of the product mix In both countries with respect to major product groups, it is less satisfactory within subgroups such as meat, poultry, clothing, and household appliances; In addition, many products Important ln one country are not found at all in the other country. With respect to health and education services, for which no suitable measure of output can be calculated, the ruble/dollar ratio was computed on the basis of the prices of inputs (wages of workershe cost of materials purchased). The ruble/dollar ratios calculated for individual products und services were then grouped into four(food, nonfood goods, consumer services, and health andnd average ruble/dollar ratios for each subcomponent were calculated by weighting the individual ratios by the shares of theln totol GBP in the United States and the USSR, respectively.5ollar ratios for the four subcomponents5ollar ratios with the use of opproprlate price indexes available from published and unpublished data of the Department of Commerce. These indexes ore those implicit in data aeries given in current and in constant dollars.**
B. Inves troent
Separate calculations were mode of5ollar ratios for three broad categories of investmentmachinery and equipment, construction, and inventories. The ratios for machinery andare baaedtudy by the RAND Corporation5 wholesale prices of moreroducer durables In the United States and the USSR. The ruble/dollar ratios for these Individual Items were grouped into several product categories (for example, agricultural machinery.
** For
* ore detailed explanation of the derivation of the ruble/ dollar ratios for consumption, see.
details and sources for these and all other price indexes5ollar ratios5ollarthe footnotes to Table, below.
explanation of the sources of data and the procedures ruble/dollar ratios for machinery and equipment, see
trucks and buses, and railroad equipment) and average ratios were computed for each category, using as weights the value of sales in each country. Tb* product categories were then grouped into four broad sectoral categories (agriculture, industry, transportation andandnd appropriate US-weighted and Soviet-weighted ratios were selected for each sector by inspection of the array of ratios for the product categories. Finally, the sectoral ratios, weighted by their shares in the total value of tbe macnlnery andcomponent ofne united States and io the USSR, were averaged to obtain the overall ratios for machinery and equipment.5 price ratios were then converted$ollar ratios by dividing them by an appropriate price index published by the Council of Economic Advisers.
The ratios for construction* are basedomparison of the cost ofoviet construction projects5 ruble prices with the cost5 dollar prices ofomparable construction projects in the United States. The size and composition of the sample of projects were limited chiefly by the data available on Soviet construction, and,esult, the projects compared arc more representative of construction in the USSR than in the United States. Tbe unweighted ratios for theairs of projectstrong central tendency formall sample, however. To obtain the overall ratios for therojects were first grouped into five categories (housing. Industrial construction, highway construction, transportation and coBBunications facultiesexcept highwaysand commercial and all othernd average ratios for each category were obtained by weighting the individual ratios by their shares ln the total value of construction in that category in the United States and in the USSR. The ratios for the five categories were then weighted by their shares ln the total value of construction In the two countries to obtain5 ratios for constructionhole. These ratios were then converted5ollar ratios by dividing them by the price index for construction implicit in Oata published by the Department of Commerce.
5 ratios for Inventories, which include both manufacturing and retail trade, were derived generally as follows; (l) separate calculations were made for manufacturing and for retailo the extent possible with available information, the total value of inventories ln each of these categories was distributed among some six to eight product groups (for example, food products, chemicals, and buildinguble/dollar ratio available from tht detailed calculation of the ratios for consumption and investment was uaaigned to eacho obtain the average ratios for manufacturing inventories, the product group ratios were aggregated
etailed description of the derivation of the ruble/dollar ratios for construction, see source rj/.
-
in two stages, using as weights, insofar as possible, the additions to inventories5 lo that group ln eachimilar procedure was used to obtain the average ratios for retail trade; (k) the ratios for inventorieshole are the averages of the ratios forand retail trade, weighted by their respective shares ln the total inventory addition5 in each country. Appropriate price indexes published by the Council of Economic Advisers were used to convert5 ratios5ollar ratios.
C. Defense
The procedure used to obtain the ruble/dollar ratios for defense differs from that used to obtain the ratios for the other ead uses of GRP. "The ratios for defense were derivedyproduct of the methods used to estimate Soviet defense expenditures in rubles and in dollars. In contrast to tbe method used for the other end uses, total defense expenditures in tbe USSR were calculated directly in rubles and in dollars for each year. In deriving ruble values for Soviet defense expenditures, known or estimated ruble prices were applied to the physical estimates summarizing available information on theof the total military structure. Forignificant amount of Information is available on the Soviet military pay system, which was applied to estimates of the Soviet order of battle to obtain estimates of personnel costs. Similar calculations were made for military procurement, facilities, and other major expenditure For most military equipment, however, Soviet prices are not knovn, and the prices for these items had to be estimated In dollar cost and then converted to rubles by estimated ruble/dollar ratios for Individual items. The dollar valuation of Soviet militarywas made by applying estimated dollar costs to the Soviet programs at the same level of detail and in the some manner as for the ruble valuations. In estimating the dollar cost, the guiding objective was to determine what it would cost toiver. Soviet program in the United Statesfor example, the cost ofank in tho United States was estimated, and not the dollar priceomparable US tanx. Some fifty major expenditure categories were considered in compiling the ruble and dollar valuations of Soviet military programs, andozen widely varying ruble/dollar ratios ore implicit in these calculatione.
The results of these calculationseries of estimates of Soviet defense expenditures for each year5 rubles and3 dollars. The resulting aggregate ruble/dollar ratio for defense increases from year to year because of the changing mix of defense activities. The Soviet mix has been shifting steadily fromcheap (in rubles) manpower to relatively expensive advanced weapons.
To calculate (JSimilar effort was undertaken but in much less detail because of the absence of tbe required data. US defense expendituresollars were divided intoajorand these categories were converted to rubles by appropriate ruble/dollar ratios obtained in the calculation of Soviet defensedescribed above. Thus the conversion of US expenditures each year to rubles reflects the US mix of major categories of defense but not the detailed US mix within each category. On this account the ruble value of US defense may be somewhat understated.
D. Administration
The ruble/dollar ratio for administration was derived as the ratio of average ft^ntmi earnings of persons employed in governmentin the two countries. In the USSR the average earnings are those of all Soviet employees in the organs of state and economicin the administrative organs of social and cooperativeand in communal economy. In the United States the average annual earnings are those of federal, state, and local government civilian employees (excluding education). Inasmuch as the ruble/dollar ratio for administration is baBed on this single comparison, the Soviet-weighted and US-weighted ratios are identical. 5 ratio5ollar ratio by dividing by an index of average annual earnings of government employees published by the Department of Commerce.
-
Table 5
FfrxHlurcp*iM of aiaai ud
a Juration (/
fal ItfdMMf Of ftOOda
for vumi' NutiiSithar bntaflu Jy1and local pjrchaaci or t
-
Total publica aun of
. aadet* ara public ccutrdctloa firJ/ Maiore/
ffrdl'.arta aor-
lnvra'.Hn'm
frdaral purehaacs of goodt and HtnHI for public Mai ll aod
HCMmdi and trrricMHfajMola andtal cara t/
Stala andor public tiaalUi'haalu,
m public canatiMctlrn far EoapIUUa and liatltEMsaailic ecaatrwfavwer and aalar wr/ Lesaeouljramlfcr IfalM-or
-and aanl:auon ml Fubllc MalUi andaalonraest (liaailma iba of tlna
A f Warn 1,
T) * TWi'rrelcaon p. il.
wj.iw
,9fo
WJ
n
1
1
*
M
*
Table 5
US: Derivation of Gross Rational Product, by End Use, la Current Dollars
Continued)
II. ftvMUrat
/
oU,.t lhan
Miliary))
fj-lpn-r. piKhaaad
OWar IMA Mfaoaa 1/
t*ard Ercxtuailon MlMR lot* aod Local atoalclap
i
Stt ciasrta of aooaa and atmfrs
lixi"ni (ltaa Iawi Jf :tniuraaa? to]
ofas*far Mftatt ty
aefacllltlra h/ lias all Hiry aaalatanc*ttata andalalia k/
. - -- - / touilaa*fhmfa S)
1. Oovsrmnil laraNaaaa af
bIhub
j. frnda mr'. in.iii
for defoaaa j/
J. a an raaltc BaaatraMta <U*
y
- -
r vmm acfaaaa 1/
J. LraaHarallm --
Kin am
S
TB
lis
_1HL_ JW "fa e Mi
it.iOl
l.Ul
6
MP
f.Ml
k.ODC
t.TO
M
no
no
337
11
t- J
" ,i
i
Continued)
fl. Expenditures on nsctiinery and equipment include gross private domestic Investment in producers' durableshare of equipment purchases in state and local purchases of goods and services for atomicther governnent purchases of equipment for civilian purposes (line IIromgovernment expenditures on atonic energy development ere allocated arbitrarily between equipnent andto the relation of producers' durablend construction ln gross private domestic investment.
i.-jiludes total new construction activity, both public ond private, less construction cfthe estlnated construction component of state and local expenditures on atonic energy developmenta,nd federal government purchases of defense production
inventories component cf investment Includes the change ln business inventories and the net acquisition bygovernment of strategic materials, %k/
) throughrom Toble 5-
Table 7
US: Gross National Product, by End Use,3 Dollars.
5
Total gross national product b/ 8
a- End-use components3 dollars uere derived by deflating the components shovn in current dollars ir., and the deflators; nre explained
in the footnotes to Table 9. DeflatorsJ1 vere derived in the same way from the saae
Department of Commerce sources.
of rounding, components may not add to the totals shovn.
foreign investment, military assistance,efense stockpiling.
3 US $
fc/
consumption b/
8
1
U
r. food goods
^
services
ducation
c/
'
9
c/
w I- <A
1 iv lu vo
?
<0 O n O t- CDC O
US fi, ^ 0 Q
T- n rn
eo 10 cu no ^
j on ii
1
e3>
a
la
l-
li C
V Ji (I
o
5
f
" -4
' S >
.
si!
.1
i g
t* o
-ri
* 0
-
e Oa0
,
j8 ::
2
II^
. i
-
II;
ill.
3 1
m
f
.
0 m
0
d
H
d
,::
s
d d
H
*
d ie
.
-
<
3
-
<
H til S:
- uS -
"
R
i
"W
St
43
ir
O
,;
us
s
CU
VJQ
-J
.:'
a, -
i
us
. I
1
<-
'
UN -1
s
Ml
On
'-'
3
'
.-'
US
us
WSC^OS <U j
r- -
r !
c
"si
* u
z r.
. -
c A
--
to
O (H
H
Bl
Ul
ll-l
"'
-9
3
s
St
Cl
ffsjo
i
3
<0
o
<n
i
IA
oi
On
O
% ft'
Ul (U
H co
a*
V)A
O 6
UlOl CO IA CO J
OJ .1 rl
g 3
5
iv* vtivo* ivU^juiiox riV'iuiv; uy kjiu out
*
6 ? 8 9 0 1 ? 3
Croup MtlOftft) product
'JS weights
USSR weights
Geometric average
Conawpfiloti
US weights
USSIt weights
Geometric average
Invect-nent
US weights
USSR weights
OtQWtric ftvernge
Defense
US weights
USSH weighs
Geometric average
Administration
a.The ruble value of US cxpendiUuca in Tableivided by the dollar value cf US ex-penditures in Table 7.
ruble value of Soviet expenditures laivided by the dollar value of US
oxnenditjres in
geometric wean of the US-weighted and Soviet-weighted ruble/doliar ratios-
d- The US-weigh ted and Soviet-weighted ruble/dollar ratios for each year are the result cf romparlng US expenditures in dollars vith US expenditures priced directly ia rubles and Soviet expenditures priced directly in rubles and dollars.
e- The US-weighted und Soviet-weighted ruble/dollar ratios for administration are the sane (oee
n Hi
t
Table 13
(Continued)
Civilian research and development
(Continued)
States k/
mm 1/
0-3
2.6
...
SSB n/
Defer.ie
5.*
0.1
5.0
;-5
USSB g/
tables
2
0
fc.T
0
1
8
3
United State* g/ 'JSSS r/
a- * af cOEinveataent and investment expenditure* forln the United ftatesrj dollars- xpenditures on education ln current dollars froaave been deflatedombined Index or wages In educationnd an index of prlcea for nondurable goods. The wag* index is baaed on the Department of Conmerce aeries of average annual earnings In public education, ln educational services not elsewhere classified, and ia coemerclal and trade schcola. ndex for nond'-irable goods la the Department of Coseerce deflator for nondurable goods. Investment expendlturea in current dollars are free. Table 6. Thesenave been converted3rt by tbe laaillclt price deflator for nev public construction ofbuildings. For the sources of the deflators uaed for nonlnveatnent as well aa investment expendlturea, see Table
expenditures for education (investment, wages, and aaterlala purchases) in rubles multiplieduble ratios derived la Table 9-
expenditures for education (lnveatmont and noninvestment) aultlplled by5ollarln Table 9.
Table 13
US and USSR: National PolicyContinued)
values of Soviet expenditures5 on wages and materials purchases in education arc estimated InThe value for vagoe5 ws =ioved forward by an Index of ereploynent In education, excluding science,value cf materials purchasesS USB moved forward by an Index estimated by thla Office from Soviet Soviet investment In education Is also estimated from budget data.
expenditures for Industrial Investment In current dollars Include expenditure on ncv plant and equipmentmining, and public utilities and construction activity ln petroleum and natural gas drilling.value ln current dollars was deflated3 pricea by the use of -unpublished separat- deflators forIn manufacturing, mining, ond public utllltieo supplied by the DepartMnt of Owmerce.
or industrial investment (machinery and equipment and construction) ln3ublo ratios derived ln Table 9-
expenditures for Industrial investment multiplied by5ollar ratios derived ln3 dollar prices was assumed to be half equipment end half construction. Although there Isway of estimating this breakdown, the division used corresponds roughly to that estimated in.
date on investment in industry (excluding Investment In pipelines) ln estimate prices was dividedand equipment and construction-Installation work (including associatedsing the relativein source on machinery and equipment and on construction-installation work5 prices by multiplying the values in estimate prices by factorsdjustments are based on data reported ln.
1. ivilian research and development expenditures in current prices were estimated by subtractingercent of federal funds spent on research ond development (the approximate magnitude of research and developmentfor defense) frcci estimated total expenditures for research and The residual values were converted3 pricesrice index for research and development estimated by this Office. J. Saviet expenditures fcr civilian research and development In rubles multiplied by3uble ratio derived in Table 9.
k. US expenditures for civilian research and development in dollars multiplied by5ollar ratio
derived ln Table 9*
1. Estimates of this Office.
"JS expcndlfores in current dollars estimated as the sum of gross new grants under assistance progrons, new credits, and other assistance through the net accurjulation of foreign currency
n. The sum of current dollar values of drawings on economic aid to Communist countries and military and economic aid to nor.-Coranunist countries as estimated by this Office. 0. ?rcci Table 7.
; . .
q. Tron
Table Ik
Previous Ruble and Dollar Estimates of Soviet and US Gross National Product, by End3
ftiiiparlBon
lar Ratio.
ublea
ViIlai--
e/
5 Ruble Retire
-
-t]
conaunptlon
sf
g/
non rood goods Cenauner service*
6
8
0
in and eofctcal'cD.
Katerlala
J.I
5.2
f/
*y
Invest.ncnl
tf
i
ery aad etjulpneniuction
Other
/
5.0
6-5
liit rat ten
dlaerepancj
grosi national product
t/
1/
Footnote* follov on
*
APPENDIX C
iterekcks
Milton, and Kravls, Irving B. An International Comparison
of National Products and the Purchasing Power of Currencies, Organization for' European Economic Cooperation,. U-
"On Measuring Productive Potential and
Relativeuarterly Journal of Economics, Aug 6l, vol LXXV, no.
, Soviet national Accounts8 CIA/RRccpariscn of Consumption ir. the USSR
and the US, U.
Abraham 8. Prices of Producera' Durables in the US and the
USSRAND Corporation,5
CIA. CIA/RRbove).
CIA. CIA/RRousehold Appliances in the USSR During the
Seven Year.. C. 7- CIA. omparison of Soviet and US Prices for
Electronic. if 8. CIAi CIA/RR, Abundant Signs of Poor Quality ln Soviet
Bwy Inctostrlsl Equipment, gJ 9> CIA. CIA/RR, Prospects for Production and Delivery
of Agricultural Equipment jr. the USSR, C.
, Rerent Developments in the Chemical Industry
of the USSR, S.
CIA/VlH, US Industrialists look at the Soviet Machine
Tool Induatry, S.
CIA/RR, ffoviet Construction Practices in Underdeveloped
Countries of the Free World, CIA. CIA/RR, The Dollar Cost of the Soviet Chemical
Ir.vestaent Program, CIA. CIA/RR, Poor Quality of Construction in the USSR
Contributes _to Future Problem, IA. CIA/RR, Hew Developments ln the Building Materials
Industry In the USSR, CIA. CIA/RR
CIA/RRonparison of Capital Investment In the US
and the. U.
CIA/RR5 Ruble-Dollar Ratios for Construction
ln the USSR and the US,
US Inccnc arid Output,. urvey of Current Business, Washington,. V.
S Income and Output,- U.urvey of Current Business, Washington,I9. U.. '
* 9
P* 2k- U.
23 n^ntal Finances.b. . 8. U.
25- S Income and Output.. U.
jurvev of Current Business. Washington,. 6. U.
S6- S Income and Output, toting,fl r m
. u.
^ of Governmental Finances. ieaKjad^itEi^. U. ' U'
U.
u.
p. Ik. u.
uTenS.BusinpSR'
29. urvey'of Current Business.ncome and Output.
if:; jui
. 8. U.
V.
. u.
Ibid.
US Income and Outp.t..
US Income and Output,. Business. Washington,.
ualnMa- Washington,. U.
bove)'.
CIA. cia/rrbove).
Conferee. Income and Output.. U
Survey of Current Business. Washington, Jul 6k, p. u.
39- Ibid.
kO. ncome and Output. u
Ibid. Survey of Current Business, Washington,'p. 30 Report Pr the. U.
Cot.
and Output..
U.
-i'V'v
U. and Output. Washington, lOpO,. u.
'J.
OOJiuncr laid.,
TKI 'I
f Current Business. Washington, Jul 6k,
>porl of the President. Jan 6k,ncome aid Output.f Current Business. Washington,utput,.
urv<
1 Current Business. Washington, Jul 6k, nd Output. Washington,. urrent Business. Washington, JU] 6k', n. $k u
-
4?. CIA. CIA/RR. U.
IDld.
Ibid.
U.
CIA. CIA/RRon of Conmwption In the USSR
and the USan 6U,. U.
CIA/RX U.
54. CIA. CIA/RR,. U.
CIA. CIA/RR. CIA. CIA/RR1 U.
CIA. CIA/RR ER U.
USSR, Central Statistical Administration. KapitaTnoye
SSR (Capital Construction in thebocov, ISC I, IT.
. U.
59- CIA. CIA/RR,l. U.
CIA. CIA/RR, 0.
USSR, Central Statistical,. U.
CIA. CIA/RR, U.
Commerce. US Income and Output,. U.
Schrooder, Gertrude. "Industrial Wage Differentials in tho USSR,"
Soviet CIA!CIA/RR, Waae and Pension Increases: The Soviet War on Poverty,. 2. C.
ClF. U.
CIA. CIA/RR, U.
Ibid.
CIA. CIA/RR ER, U.
Ibid.
CIA. CIA/RR, 0.
CIA. CIA/RR,. 0.
US Congress, Joint Economic Committee. Annual Economic Indicators
for the USSR, 0.
Commerce. Survey of Current business, Washington,. U..
CIA. CIA/nR ER U.
nc. Struktura kapital'nyhh
vlozhcnlySnA (Structure of Capital Investment in the USSR and the U.
Central Statistical Administration. Kapital'noye
SSR (Capital Construction in the- Cosweroe. Statistical Abstract of the unitedashington,
80. . U.
,. u.
-
Original document.
Comment about this article, ask questions, or add new information about this topic: