Economic Intelligence Memorandum
THE SOVIET PRICE STRUCTURE AND THE RATIO OF MILITARY EXPENDITURES TO GROSS NATIONAL PRODUCT
CIA HISTORICAL REVIEW PROGRAM RELEASE IN8
CENTRAL INTELLIGENCE AGENCY Office of Research and Reports
Economic Intelligence Memorandum
THE SOVIET PRICE STRUCTURE AND THE RATIO OF MILITARY EXPENDITURES TO GROSS NATIONAL PRODUCT
Thu material contains InformationNaufaal Defense o/ the Unitedythe mcanlnjt of the esplcmn. USC. Stat THrhich Inf iohLbllea* bv law
CENTRAL INTELLIGENCE AGENCY Office of Research and Reports
H. Concept of the Coat or Military Pro gran
III. Objections to Estimates Based on Soviet Price
IV. Evaluation of the Ratio of Military Expenditures to
Expenditures on Final Product at Adjusted
of Soviet Military Expenditures to GHP at Ad-
- Ill -
THE SOVIET PRICE STRUCTURE AND THE RATIO OF MILITARY EXPENDITURES TO CROSS NATIONAL PRODUCT*
It is frequently stated that existing estimates of the cost of Soviet military expenditures, calculated as the ratio of militaryto gross national productould increaae substantially (perhaps double) if Soviet prices reflected costs in the Western sense. This memorandum shows that very large hypothetical changes tn price relationships among categories of final goods andhosen specifically for their tendency to increase the ratio of militaryto GHP, would not affect the ratio by more than two or three percentage points.
* The estimates and conclusions in this memorandum represent the beat Judgment of this Office as of
** The term burden is often used to describe this ratio. In this memorandum the term cost or thc phrase relative allocation of resources will be used so as to avoid any implication that the subjective effect of military expenditures (the extent toopulation feels that it Is burdened) can be measured cither by auch ratios or by thevalue of military expenditures.
he difference between the prices paid by civilian and military purchasers for alallar nilitary and nonmilitary products Is believed to be negligible (see the first footnote on p.elow).
Therefore, although itough measure of cost, the ratio of military outlays to GNPonvenient and useful approximation of the proportionountry's resources devoted to the military sector. It must be eophasized, however, that military expenditures do notmilitary power ln any absolute sense, und, of course, neither do ratios of military expenditures to GNP.
Attempts have been made repeatedly toethod forthe cost of tho Soviet military program, but the estimates that have been derived thus far have been objected to on the ground that Soviet prices tend to understate the "true- magnitude of military Thia memorandum discusses the concepts underlying one measure of this kindthe ratio of military expenditures for final goods and uervlceu to GNPand examines its applicability to Soviet data. The reliability of this measure is then evaluated by comparing hypothetical prices for goods and services with actual prices ln order to test the dependence of this ratio on the price structure thatthe existing estimates of Soviet GNP.
11- Concept of the Cost of Military Programs
The coatilitary program can be defined aa the share of the national product claimed by the program.* For Internationalthia cost is often measured in terms of the ratio of militaryfor goods and services to GNP (the value of the totalof goods and* This memorandum examines the validity of this ratio. The use of this ratio Implicitly assumes that the value of the product in the military sector can be compared with the value of the product in the civilian sectorhat Is, the price of any product equals its opportunity cost (the value of an alternative product that is sacrificed by producing an additional unitarticular product).
* Military programs have Implications for an economy other than the donlal of the alternative products that could be produced with resources allocated to the military sector. For example, by providing purchasing power but neither the goods for which lt can be spent nor the means of producing goods, outlays for these military programs create Inflationary pressures. For tbe purpose of this memorandum, however, these andproblems Involved In supporting military programs axe not relevant and will not be pursued further.
For serially numbered source references, see the More precisely, the price of the final product must equal the marginal coat, and the marginal cost must /Footnote continued on p. Kj
The validity of the assumption referred to above depends ono which the prices paid for final products both in and outdefense sector actually measure the relative desirability ofand services us established by the interaction of consumersdictateslanning authorityhe extent to whichreflect the full costs of the resources used in producingand services. When satisfied, these conditions implyroduct equals the value of the product .foregone at
The Ideal conditions necessary for equality of prices andcosts are not satisfied in any economy, nor are they likely ever to be met. In all countries the compilations of GNP are subject to the criticism tbat they do not reflect fully and accurately the relative economic coBt of producing tbe array of final goods and services. The gap between the ideal and the real Is greater lo the USSR than lneconomies, for prices in tho USSR vary widely and unevenly from opportunity costs, nevertheless, the difference ls one of degree, and useful comparisons of GNP aggregates may appropriately be made within or between economies. Such comparisons are made regularly by several organizationshe UN, NATO, and the ECE, for exampleand are valuable If appropriate caution ia observed in interpreting them.
Before proceeding to the specific objections made to estimates based on Soviet price relationships, two additional points must be made. It ls not legitimate to argueransfer of largeof resources from the military sector to the civilian sector would necessarily increase the value of civilian output as much as it reduces the value of military output. Even if prices equal opportunity costs before the transfer is made, this equality dependsiven allocation of resources and holds true only for small changes in this allocation. Large-Bcale reallocations of resources would change price relationships among final products because of differences infunctions and demand schedules between the civilian and the military sectors.
In addition, it must be emphasized that the cost ofiven country can be expressed validly only in termo of that country's own prices. Some estimates of the cost for the USSR seem to have been derived from dollar valuations of both Sovietexpenditures and Soviet GNP and have resulted in ratios of Soviet military expenditures to Soviet GNP that are considerably higher than the equivalent ratio based on ruble value3. Because dollar valuations of Soviet GNP reflect the cost structure of production in the US, and not in the USSR, it would be only accidental if the valuations measured the economic coat that defense expenditures Impose on the Soviet
equal the prices that have to be paid for the quantities of all the factors of production required to produce an additional unit of output. If the prices paid to all factors of production are equalizedthe economy through some sort of bidding process, the transfer of small quantities of factors from one employment to another subtracts as much from the value of the total output as it adds to it, and prices arc equal to opportunity costs.
Differing relative costs reflecting different resource endowments and technologies and differing demand /footnote continued on p. Jy
m' o Estimates Based on Soviet Price Relationship.!
A number of economists have argued chat thc cost of Soviet mllitarv expenditures cannot be estimated, because of shortcomings io Soviet prices and in other available data. These writers maintain that then calculating CNP, and the division of CN? by end use, do not meet the requirement that the value of output ahould equal the costs of producing output throughout the/ As stated above, this idealized requirement is not fully met ln any economy, but there are reasons for believing that the failure isstriking ln the case of the USSR. The question le whether these shortcomings seriously affect the measurement of the cost of the Soviet military effort, within the limits of the purpose of such
It is beyond the scope of thia memorandum to inquire into whatpricing system might imply for tho valuation of Sovietcan be applied properly to
ahtaSLtT*: !summarizes instead some of the principal objections to existing calculations of Sovietbjections that have been made on the grounds that the calculations do notalid picture of the relative costs of production. This memoi-andua also at-
frecttaMng these objectionsmight have on existing measurements of the ratio ofto
ead to significant
Mf-erenees in their price structures. That the Htrue" Soviet price
* These criticisms refer to distortions in the price structuretortions wiStn^e8
b^fln I* 1, 1
based on available evidence, ia that price discrimination in favor of military purchases ls not sufficiently Important to be considered,iscussion of this problem, see source jj.
One objection centers on the distortions introduced by indirect taxes and subsidies when Soviet GNP ia calculated in market pricesverstatement of the prices of consumer goods relative to the prices of producer goods. Move describes the effect of the turnover tax, the primary offender, on the price structure in the following manner:
Ic is broadly eue to say that producer goods prices areto. and usually intendedovei, their cost of production,nauiner goods prices should be ia principle related to Icraeod; the difference between tbe cost of production and this "supply and demand" price equals turnover tax. plusoulier eatear)be high prices of consuacr goods enable tbe state toad orher "noo-rooswap-uoo" cipeaJiture while restrietiag effective demand to tbe com-snaer goods wfcicb it decides to stake available. 4/
This distortion can be remedied ln part by adjusting CUT' lor turnover tax and subsidies, provided that those taxes and subsidies can beby industry or product.
Although calculations of GNP can be adjusted to representas the USSR computesundamental problem remainsto vhlch Soviet prices do not include the full coats of There is general agreement that amortization of capitalin calculating the average cost of Soviet production.again the impact of such an understatement would affectindustries and, therefore, the producer goodsmost heavily. It is argued further that prices inarger profit element than in producor goods6/ As lt la hardly likely thatituationdifferences ln the quality of management or the degree oflarger profit in consumer gooda also would cause an understatement
of the prices of producer goods ln relation to the prices of consumer goods.*
Perhaps the most persuasive objection to estimates baaed oo Soviet pricing practices consists of the fact that almost no rent is charged for the use of land or interest for the use of investment funds. Jj The existence of this practice is an additional reason for believing that the prices of producor goods should be raised relative to those for consumer goods to bring prices into line with the average costs of production.
The reasoning above, restricted as it ia primarily to capital andassumes that in the labor market the prices paid to factors tend to be equal In alternative employments. Although this assumption would have been unrealistic for some periods of Soviet history, for the period since Stalin'a death it probably would not be too unreasonable.
* arket economy, one would normallyransfer ofout of the production of producer goods into tbe production of consumer goods shouldituation exist, thus forcing the prices of consumer goods down and leadingise in the prices of producer goods.
There is now much Icse direct interference with labor mobility, and lt is assumed, therefore, that relative wages and the labor market are so close to equilibrium that they can be Ignored for the purposes of this memorandum.
In spite of the validity of the objection outlined above, there is no practical basis for adjusting estimates of Soviet ONP (stated in market prices) to meet them (although some distorting effects can be removed by subtracting indirect taxes and adding subsidies). For this reason, an indirect approach must be used to evaluate declarations such as the. It seems likely that the figure ofercent of CNPdrastically underestimates the proportion of real resources used for national security/
TV. Evaluation of the Ratio of Military Expenditures to GNP
Estimates of Soviet GNP0ave been constructed, both in established prices and adjusted as indicated above. Sj Military expendituresarger share of GNP when GNP is adjusted for indirect taxes and subsidies than when ONP Is calculated in established prices because the military sectormaller proportion of those goods subject to high rates of indirect taxation. The followingcompares0) the ratios of military expenditures to GNP on an established price basis with those calculated on an adjusted factor cost basis:
BillIoa Current Rubles'
OW ot established prices
Facio of ollitary expenditures
to CHP tM>
GXP at adjusted factor
Ratio of ailitary
* The use of any ruble-dollar ratio in this memorandum would beInasmucheparate ruble-dollar ratio exists for each commodity and the application of any ratio would distort thebetween defense expenditures and GNP (see the third paragraph on p. k, above). All values have been rounded to theillion rubles.
to GXP HO.U)
The estimates of GNP include breakdowns by end useonsumption. Investment, defense, administration, and other. By rearranging the accounts, lt Is possible toough estimate of five major types of final products entering Into these end usesfood, machinery and equipment, constructloo, materials and miscellaneous manufactured goods, and personal services and utilities.* For the purpose of this memorandum. In tasting tbe effect of further relative price changes, it is sufficient to work with this fivefold breakdown in the military sector and In the total economy, as shown In*
A ranking of the major types of final product according to this relative importance in military expenditures compared with theirImportance in GNP forears shows the following:
Percent of Military
Rxpendltures Percent of GNP
Machinery and equipment Personal services and utilities Materials and miscellaneous
manufactured goods Food
of the differences In the distribution of final output between the military and civilian sectors, it is evident that any increase In the prices of the final output of machinery and equipment, personal services and utilities, and materials and miscellaneous manufactured goods would increase the share of military expenditures in GNP. It la this sort of adjustment that critics have in mind when they Insist that existing estimates of the proportion of Soviet resources channeled Into military programs are grossly understated.
The appropriation and capital accounts of the household,enterprise, nonagrlculturel enterprise, and government sectors were allocated according to type of final product (consumption and Capital repairs and new fixed investment were allocated to construction and machinery and equipment2 toasis. for military procurement and operation were allocated by end product according to estimates compllod In this Office. The adjustment necessary to eliminate the effect of indirect taxes and subsidies was prorated among the end uses by type of final product following the guidelines established lnf Soviet Nationalollows on p. 9-
The importance of this alleged understatement can be tested ratber readily. Relative prices of final output can be manipulated bya set of hypothetical price Increases in the product sectors that vould Increase the ratio of military expenditures to GNP. Thceffect of such an adjustment of tills ratio provides an Insight as to the dependence of tbe ratios on the set of price relations embodied in existing estimates of Soviet GRP.
Revisions of the value of final output along these lines show that it is unlikely that the ration would be affected materially. f Tableives the ratio of military expenditures to adjusted GNP. If the prices of machinery and equipment were doubled, this ratio would Increase about two points, as shown ln item 2. hange of this* magnitude can hardly bo considered large, in view of the gross nature of the measure.
None of the objections cited above has suggested that the return to personal services, by far the largest part of the services and utilities sector, is underpriced relative to other components of GNP. The possibility could be raised, however, that on increase in the value of the services and utilities sector would raise the ratio of military expenditure to GRP. Such an increase would seen unjustified. Because of the relatively free mobility of labor, the returns to personal services in this sector are related to the wages paid In other sectors. Therefore, on upward revision of the wage component of services and utilities mightike Increase In the wage component of each of the other product sectors, pertly negating the effect of anin tho value of services and utilities. Moreover, the average pay and allowances and payment ln kind (except housing allowances) for personnel of the Ministry of Defense and thc Militarized Security Porces exceeds the average annual earnings of workers and employees in state enterprises, organizations, ond institutions in each ofears. Thus, before any general adjustment In this sector could be made the personnel costs of the military sector probably should be revised downward to reflect what military personnel could earn In the civilian economy. This procedure would reduce the difference in the importance of personal earviceB and utilities as between the military sector and tho economyhole and vould dampen the Impact of any generalof prices in this sector.
* ollows on p. U.
The category "Materials and miscellaneous manufactured goods"ide variety of final output. Because production of this outputls quiteargeercent) in the price of this output is incorporated, and the result of this change, added to the effect of the posited changeercent In the prices
Ratio of Soviet Military Expenditures to GNP at Adjusted Factor0
similarly. Price increases in this sector, however, would tend tothe original, ratio between military expenditures aod GBP. For example, if prices were raisedercent in construction and if the impact of thie change were added to the other changes, the ratio would fall in every one of the years as illustrated Inf Tablehe larger the increase in prices of construction output, the greater is the decline in the ratio.
Also of interest is the fact that the adjuated ratios are not very responsive to changes in the composition of defense expenditures durina
SeniSIJJ"' Although machinery and equipment
accounteduch larger percentage of military outlayshan0his change in composition'had little impacVon the effects produced by adjusting the prices of machinery and equipment.
The effect of other combinations of price adjuatmenta could be tested, but the adjustments already discussed demonstrate themodest effect that they would have on the meaaurement of the economic cost of military expenditures. Although the breakdown of final output within the military sector and GNP is admittedly crude, it would take rather drastic rearrangement of expenditures and prices to push the ratio of military outlays to GNP much higher than that shown inf Table 2. The reason Is that the military sector does not differ enough from the economyhole in its pattern of demand for final output of the goods and services. The situation would be quite different if the focus were shifted to the ratio of investment to GNP. Here hypothetical increasesercent inand equipment prices andercent in construction prices change^the share of Investment in GNP6 fromercent to 37
* This conclusion is dependent on the substantial validity of the Judgment that price discrimination in favor of military purchases ia not sufficiently important to be considered (see the first footnote
The conclusion to be drawn from the analysis, then, ia that the cost of the Soviet military program, as measured by the ratio ofexpenditures to GNP, is not seriously misstated, providedare made for indirect taxes and subsidies.* Although the ratioaeoaure of cost, it satisfies reasonable" demands Placed on it. Military expenditures do not represent military power in any absolute sense, and, of course, neither do ratios of military expenditures to GNP. It would be irrational for one country to design its defense policy with reference to the ratio of defenseto GNP in another country. The usefulness of the ratio appears to be limited toeasonable approximation of the proportionountry's resources devoted to the military sector.
Evaluations, following Che classification entry and designatedave the following significance:
Confirmed by other sources
Cannot be Judged
"Documentary" refero to original documents of foreignand organizations; copies or translations of such documentstaff officer; or information extracted from such documentstaff officer, all of which may carry the field evaluation
Evaluations not otherwise designated are those appearing on the cited document; those designated "RR" are by the author of this No "RR" evaluation is given when the author agrees with the evaluation on the cited document.
Except for CIA finished Intelligence, all sources arc evaluatednless otherwise indicated.
1. Hoog, Malcolm W. "Economic Problems ofournal of Political Economy,o . U.
Hitch, Charles J. and McKean, Roland B. The Economics of Defense in the Nuclear Age, U.
Gilbert, Milton and associates. Comparative National Products and Price Levels, Paris, nd- U.
Security, New. kk. U. Eval. RRovo, Alec. "Some Notes on Soviet National Income Statistics,"
Soviet Studies, volo- U. Eval. RRootias, John. "Price-Setting Problems in the Polish Economy,"
Journal of Political Economy,o k,.
U. Eval. rr 2.
3- Montlas, John. "Rational Price and Marginal Costs in Economies of the Sovietoviet Studies, volo k,. u.
Dobb, Maurice.omment on the Discussion About Price Policy,"
Soviet Studies, volol*2. u.. "The Adjusted Factor Cost Standard ot Measuring National Income: Comment," Soviet Studies, volo. U. h. Nove,it.. U.
op. cit.- U.
CIA.,0 Soviet Budget, U.
. op. cit.bove).
7- CIA. CIA/rr ER U.
Turgeon, Lynn. "Cost-Price Relationships in Basic Industries During the Planningoviet Studies, volo U.
6. Schlesinger,lt. U. Eval. RR CIA. ORR, Soviet Nationalunpublished). U.
- Ik -Original document.