PROVISIONAL INTELLIGENCE REPORT
BODTJCmm Wi THE USSR
OA HISTORICAL REVIEW PROGRAM RELEASE AS SANITIZED
CENTRAL INTELLIGENCE AGENCY
PROVISIONAL INTELLIGENCE REPORT
BCfl PRODUCTIVITY IN IHE USSR
The data and conclusions contained in this report do not necessarily represent the final position of ORR and should be regarded as provisional only and subject to revision. Concents and data which may be available to the user arc solicited.
CENTRAL INTELLIGENCE AGENCY
Office of Research and Reports
I. Total 2
Scope and CharactoriBtica of the 2
War acd Postwar 3
of tho Fourth Five Year Plan
of the Fifth Five Year plan
Affecting Soviet Labor Productivity and Its
II. Individual Economic
*. Ferrous Metallurgy
Machine Building Industry
ppendix A. .
Appendix B. Gaps In 3
Appendix C. Sources and Evaluation of Sources. 35
of Productivity in Selected Industries in the
in the USSR. .
of Production, Productivity, and Employment
In the0 and
3> Indexes of Production In Selected Industries
.In the USSR,
*. Comparison of Productivity of Industrial Sector
in the USSR and in tho0 and
6. Annual Output per Worker in Coal Mining In thc USSR,
Annual Output per Worker in the Peat Industry
in the USSR, 1*
8. Index of Output per Worker in the Petroleum Industry
9- Indexes of Output per Worker in the Ferrous Metallurgy
Industry in tne USSR, 16
of Output per Worker In the Cotton Textiles
Industry in the USSR, 19
of Labor Productivity in Machine and Instrument
Construction lo the USSR, 21
of Output per Worker in State Bearings Plant Ko. 1,
per Worker and Index of Output per Worker
in Railroad Transport lo the USSR, 25
lft. Index of Output per Worker ln Construction in the USSR,
of Production in Selected Industries in the USSR,
of Industrial Production In the USSR,
Selected Yoars, 30
Figure 1. Soviet Official Indexes of Industrial Production,
Productivity, and Employment
Figure 2. Soviet Official and CIA Indexes of Industrial
INDUSTRIAL* LABOR PRODUCTIVITY. IN THE USSR**
Increasing the productivity of industrial labor represents one of the major economic objectives of the Soviet government, the Party, and the individual industries. In the postwarighnvestment, improved technology, better management, and greater skill on the part of workers have been reflected in striking advances in output per man year in industry. Soviet official indexes shew thati8 industrial labor productivity hud recovered from the effects of World War II- 80 the Soviet index based0 continued to advance rapidly, movingon increase ofercentears- Bie Fifth Five Yeara further gain5 ofercent On the basis of reports of attainments for theears of the Plan period, however, it appears that the original goals will not be met and that thc Soviet index of industrial labor productivity5 will increase only aboutercent
There is considerable disagreement as to thc exact meaning of Soviet.figures on productivity, which may overstate the increases achieved and should be used with reservation in international Nevertheless, it is believed that published Soviet figures are useful in giving an approximation of trends in industrial laborin the USSR.
* Industrial in this report refers not only to the manufacturing and extractive sectors but also to transportation, construction,and other activities which support production. Inhe general index of productivity covers, only the manufacturing andsectors, while in Section II construction and rail transport indexes arc included.
** Hie estimates and conclusions contained in this report represent the best judgment of the responsible analyst asft-
Comparlsonii among industries indicate that tho progress has been uneven. Those industries exhibiting the earliest and Boot persistent postwar gains included the metals and Machinery industries, which wer favored both as producers of capital goods and as suppliers ofond items. Progress in the extractive industries and in consume goods manufactures was slower.
I. Total Trendu.
1. Scope and Characteristics of the Data.
Labor productivity is defined in this report as the outputroduct per unit of labor input.* Ho atteopt is made to measure separately the effects of changes ln capital Investment, technology
hC ^labor- ratberlhe
combined effect of these factors of production on labor productivity together with some comment on their general trends.
A productivity Index may be considered toraction, thateasure of production dividedeasure of employment Itherefore affected by all the errors which may occur In either the measure of production or the measure of employment selected. Inaggregateoreover, serious statistical biases may be Introduced by the technique used for weighting component parts in order to buildepresentative total. This problem is most serious in the case of measures of production, since output is frequentlyin different units which must be weighted by value or some other common measure before they can bc added together. Those
* The labor input unit used throughout, unless otherwise stated, is the- man-year, which Is treated as synonymous with persons employed. This ir. not ao precise as the use of man-days or man-hours, but laborreferable calculated from available
" Aggregate Indexes are those which combine the trends ofve- total trend, for example, combining trends in the productionumber of types cf machines Into an over-allindex or the combinationumber or individual Industry indexesotal industrial measure- Both of these levels of aggregation underlie some of thc series presented in this report.
difficulties in computing satisfactory measures of total laborhave given rise to extended discussion both by Sovietand by the Western economy who atu-mpt to use for this reason, this report does not attempt toof the pitfalls for the unwary user of productivity statisticsaumuarizes thc principal criticisms in the(see Appendix
Although thc reader Is warned by theapranh that productivity measures cannot be Interpreted as presentinT^ exact picture, the trends are so pronounced that useful conclusions may bo arrived at even from thc crude materials at hand.
they can be related, have been selected for analysis partly because differences in tbe structures of prewar and postwar economics make cc^axisons deceptive and partly because there arcrserious technical
Z ZLIt may be said, however, thatthe expansion of physical volume warrants the statement thatperiod was one of substantial achievement in production andagreed on. Tablehows advances in volume ofQlte^
Sooe of the newer industries probably gained even more rapidly in-the prewar period. Soviet claims of increases inof machineryT yield an estimateercent8
information on employment would suggest rapid Increases in productivity.
rU pcrlodtoe Soviet official index
of productivity in large-scale Industry advancedercent An has been indicated, however, this wus basedroduction index which was subject to such Inflationary biases that it is generally di"
2- War and Postwar Changes.
Postwar changes In thc Soviet system of pricing industrialessening of the relative Importance of new products included
' ollows on p. k.
Indexes of Productivity a/ In Selected Industries In the USSR
In the official index should have eliminated,onsiderable degree, the biases in the prewar production index and should make appraisal of the Soviet claims easier for current years.**
Scarcity of firm data on physical volume of production in postwar years cakes It difficult to secure independent checks on the extent to which this inflationary price bias bar. been eliminated froa tbe official Index by comparing It vith indexes based on physical volume. Tjo such efforts at estimating production from physclal(weighted in proportion to payrolls) have been made, one by Donald Hodgmaa and one by CIA/ORR. The methodological appendixtechnical problems involved ln such comparieonG.
* compares thc Ifodgoon Index and the CIA index of Industrial production with the Soviet official index. The second nactlon of the table compares indexes- or productivity derived from these production indexes by dividing them by the some index ofemployment.
It will be noted that throughout the period all of the indexes register substantial increases ln produetion. Indicating recovery to
* Footnote references In arable numerals ore to sources listed in Appendix C. ** See Appendix A, Section ollows on p. 5.
- ft -
Indexes of Production, Productivity, and Employment in thc0
series93 computed by dividing official production series by official productivity series; employment*5 projected on the assumption of smaller increases inemployment in these years than in tbe.d-e.
f. Projected on the assumption that thc Fifth Five Year Planpercent increase in production would be fulfilled.
0 level by about the beginning8teady rise The officially announced increases in production andarc consistently above those computed from the ORlt and Hodgman estimates, but there is no scientific method of determining whichgives the "truer" picture of Industrial expansion.
of the Fourth Five Year Plan
Soviet production and productivity suffered cataclysmic declines during the war except in those defense industries
located sai'ely beyond tbe area of war damage, it is remarkable that the economy recovered0 levels Thisestimony to the ruthlcssness of the drive of Stalin to rehabilitate thc productiveand in part to the willingness of the workers, at least temporarily, to cooperate in spite of thc lack of substantialin their level of living.
Statistically thc0 is characterizedise in the official index of productionorcent of0 level and of productivityercent of0 level (see The CIA indexrowth ofercent in production andercent in productivity.
Comparison of the production increases in various industries00 indicating the concentration on heavy industry is given In Table 3-
Indexes of Production in Selected Industries In the USSR 6/
Fabricated Metals Defense Chemicals Nonferrous Metals Electric Power Manufactured Consumer Goods Food Products Forest Products
b- Period of the? Firth Five Year.
Continuing the trends of theears, thefor the Fifth Five Year Plan calledurther increase in production ofercent and of productivity ofercent,onsequent increase in employment of3 percent. It was pointed out by ORR at the time jj that in view of the substantial increase in the population in the working ages this employment in-
crease could easily be exceeded if the situation required it. During theears of the period the economy followed the planned rate of expansion closely as production increasedercentercent, an annual average of 1ft percent in productionercent in productivity. 2owever, the trend changed. Production increased onlyercent,andercent- (See) Bie slackening was due in some meaoure to thc disorganization following the death of Stalin and in part to the beginningseliberate shift from the emphasis onindustries to the development of low-productivity consumer goods production.
fact that thc rate of increase in productivitymore pronouncedly than the rate of increase in production was attributable to the abnormally large increase in employment. Whereas it was originally planned to increase industrial employment3 percentt is estimated that employment exceeded this level by Several factors caused this rapid increase. Among them were the release of large numbers of 'forcedomewhatpolicy of discharges from the armed services, andetention of larger numbers of women than originally contemplated. For the first part of tbe year, at least, transfers from farm to industry continued.
On the assumption that thenitiated3 tothe "new course" will continueftt isthat the goalear increase ofercent in production will be practically attained. Owing partly to tbe more than planned increase in employment and partly to developoent of consumer goods production at an accelerated rate, productivity will not attain the planned goal ofercent increase ndicates that, according to Soviet announcementsincreased onlyercent0 instead of the plannedercent. With slower increases in Industrial production,increases will lag still further behind the planned rate. The projections of the official production increases divided by the CIA estimated employment increases will5 productivity indexercent0 instead of theercent, when calculated from thc official production index, andercent when calculated from the CIA production index. (Seend
* ollows* P.bove and following p. Q, respectively.
c- Future Prospects.
When the trends of production and productivity are plotted as in Figureoticeable flattening out of the rates of Increase appears after the sharp riseO This deceleration of the rate of expansion ia characteristic of economieseriod of rapid growth when previous gains must be consolidated and digested and when growth has attainedevel that further percentages offrom the large base arc more difficult than gainsmall base.
It would be hazardous to extrapolate this slackening growth trend0 or even7 hy mathematical formulae. There Is reason to believe, however, that thc favorable conditions operating during the Fourth Five Year Plan will not recur, nor does it soea likely that the Soviet people,ew years of improvement In their level of living, can be "weaned" froa the consumer benefits and forced to return to cxtrese emphasis on heavy Industry without morale difficulties which would lower productivity. Thus continued increase In the proportion of workers assigned to the low-productivity industries wouldrake on the rate of expansion of total industrialunless capital Investment In low-productivity industries lo sharply increased. On the other hand, some previously favorable factors will continue to militate against toorop in the annualin productivity. Among these are expanding programs for technical training at all levelson-the-job, vocational high school and technical university. It Is possible also that concentration of know-how" on the consumer industries will produce Improvements In physical plant, technology, and management In this sector comparabla to post gains in the heavy lnduotry sector.
On balance, therefore. It would appear reasonable toradual decline in the rate of Improvement In productivity from thoccording to the official Index) ofercent per year.
On the basis of-some highly speculative reasoning, Calenson hou arrived at an estimate (shown inf the relationshipproductivity trends in tho USSR and in the US.
* ollows on p. 9.
Comparison of Productivity of Industrial Labor in thc USSR and in the00
On Galenson's assumptions, it would appear that Sovietwill hardly reach two-thirds of that in the US but that thc ratio will gradually become more favorable to the USSR.
3- Factors Affecting Soviet Labor Productivity and Its Measurement.
Some conclusions may be drawn concerning Soviet efforts tolabor productivity, but the evaluation of the success of their efforts and the impact of tbe various factors would require detailed analysis' beyond the scope of this.report.
Doubts have been expressed concerning Soviet claims ofincreases because of the depressing effects of the destruction which occurred during World War II. However, an analysis of postwar changes in equipment and the utilization of equipment in blast and open hearth furnaces points out that the reconstruction of damagedIncluded modernization which would foster the growth of8/ It seems more than likely^that reconstruction in other industries also entailed modernization. This would facilitateincreases, perhsps not always to the level of Soviet claims.
Takenrude guide, the increased availability of equipment per worker should contribute to increased output per worker, although the relationship can not be measured. 0 the amount of technical
equipment* available per Soviet worker was approximatelyercent overiO level, and the electricalasercent overQ level. a, electrification per worker had increased an additionalercent toercent*
With the passage of time, the age and sex structure of thc labor force is becoming core normal, with less dependence on lower productivity under- and over-age groups. Soviet emphasis in theperiod on improved levels of training should also be contributing to Increases in output per man as the relatively large numbers of new workers added to the Industrial labor force8 use competence they have gained through experience and on-the-job The proportion of personnel with higher and secondary technical training has also increased. 0 the number of such personnel had increasedercentompared with an increase ofercent in total nonagrlcultural he former are expected to increase by an additionalercent and the latter by approximately It is also reasonable to suppose that the quality of training has Improved.
The impact of these factors varies from industry to industry. It was first felt in those heavy industries where the investments Were first made. The impact should be beginning in consumer goods Industries in which new plants have recently begun to operate and labor training plans have boon expanded.
1A : The effect of management policies on labor productivity is even less tangible than the other factors. There is, however, no question that Soviet labor control policies are aimed at increasing labor Although both recurrent agitation for productivityand incentive provisions are utilized, the effects may be weakened by indifference or other morale factors. Management has also been criticized for failure to utilize available equipment and labor Increasing emphasis is placed on improvedof work and the constant-flow method of production as sources of higher labor Improvements in transport andand thus in supply, woulS contribute to productivity increases through elimination of work stoppages and erratic production.
The technical equipment index was calculated from5 planned increase0/
Estimatedpeech by Saburov citing an increase ofercent,
evidence. II. Individual Economic Sectors.
This section describes the trends ln some of the principaland economic sectors. The level of productivity in these categories3 as compared0 (abstracted froa tbe sector sections) is shown In Table
Indexes of Labor Productivity In Selected Economic Sectors
Machinery and Instrument
Iron and Steel Chemicals
Iron Ore Mining
Indexes of Labor Productivity in Selected Economic Sectors
in the USSR3 (Continued)
a.Data abstracted frooU.
Die extent to which the producer goods manufactures have been favored is immediately apparent from this list. This has taken the form of giving.these groups highest priority in the assignment ofand skilled personnel. The construction and transportation cate-"gorles also show substantial advances- llext in order in the industrial categories are the consumer goods manufactures. Unfortunately textiles is the only group for which official figures ore available, but scattered evidenceimilar trend in food processing. With the exception of petroleum extraction the slowest progress is shown by the extractive group.
Analysis of the changes in production goals planned to build up the consumer goods industries indicates that the differential in production will be considerably- "Die effect this will have on productivity will depend on the ratios between capital and laborthat are worked out.
1. Coal Industry.
Hie study of output per man in the Soviet coal mininghas been relatively easy in the past because of theof product and relative availability of data. It is possible to construct several time series for output per worker from official sources. Bie variations arise largely from the use of employment
data referring to differents to all workers, production
workers, or underground workers.
The great need for rehabilitation of mines in the Donets Basin has impeded the recovery of the coal industry in that area The increased proportion of production now furnished by fields in the eastern regions, where productivity is considerably higher because of the nature of the coal seams and the degree of mechanization, hasthe lower productivity in the Donets
Increased mechanization also contributed to increasing output per man. l9 the cutting and breaking up of coal wasercent mechanized; the extraction of coal from the working face,ercent* and the loading of coal into freight6*
utput per man per year had surpassed the prewar level byoercent, reachingetric tonsetric tone. (See
2. Peat industry.
^viet peat industry was scheduled in
the Fourth Five Year Plan toetric tons per yearctual output per man0etric tons,er-
tpnt^ creasedthe <- ate3 output perons Annual output per worker in the peat industry inis shown in Table IK
* Details on selected mines ore given in source ollows on p.
*** ollows on pT il,.
Annual Output per Worker In Coal Mining in the
CIA Estimate Year (Metric Tons)
Galenson (Metric Tons)
Annual Output per Worker in the Peat Industry In the
Footnotes forollow on
Annual Output per Worker In thc Peat Induatry in the USSK 2jJ a/
The data may apply only to Glavtorf of the Ministry of Electric Power Stations, b. At annual rate of increaseercent.
3- Petroleum Industry.
s shown in Tableoutput per man ln the Soviet petroleum Industry had regained the prewar level. Assuming noinover8 level of output perhe index3 Indicates that output per manmetric tons of petrolcum-per year,4 metric tons of petroleum and cas cc. Mned.
Many instances of increases In labor productivity are cited as resulting from Improved organization of labor and
'i. Ferrous Metallurgy Industry.
Plan results indicate that labor productivity ln the iron and nteel industry exceededevel'byercent0 andercent Labor productivity in blast and steel furnaces increasedore rapid It appears that theTor the industryhole was lower because of low rates of Increase in productivity ln Iron ore mining, and perhaps in casting
' Table ti follows on
Index of Output per Worker in the Petroleum Industry in the
on the basis of tbe previous year.
and rolling. Changes in labor productivity In the industry are shown in Table 9.
Indexes of Output per Worker in the Ferrous Metallurgy Industry
in the USSR
Pig Iron Steel
Vear Industry Total Ore- Mining Sneltlng
Indexes Of Output per Worker in tbe Ferrous Metallurgy Industry
in the USSK
Projected at the same rate as Oil metallurgy.
b. Estimated on basis Of previous years-
a. Iron Ore Mining.
Iron ore output per worker was reportedetric tons per year. *j/ 0 productivity leveletric tons per worker per year may now obtain and may beto remain unchanged, even with improved techniques, because of the declining share of open pits in total Increased dependence on the utilization of poor-grade, ores vlll require onin concentrating operations, and to that extent lower Soviet productivity^ in terms of ore ready for use in the blast/
bl. Pig Iron Smelting.
7 thereoviet figures for metric tons per worker perndjj, the latter presumably based on thc smaller figure of workers directly employed. 1 the productivity of labor in blast and steel furnaces was reported to have increasedercent of0/ Even assuming noin productivity03 output per worker is indicatedetric tonsetric tons."
* 7 output per man is projected by the index in Table 7-
Contributing to the increase in labor productivity, the utilization of areu of blast furnaces hod increased0 bvxr-cent. ftg/
utput per worker in open-hearth shops0 metric tons or steel per Using the same projection as above, steel smelted per worker In open-hearth shops3 would approximate oretric tons per year.
In open-hearth shops, the utilization or equipment had also increased, so that the removal of steel per square meter of hearth area exceeded0 level0 by
d. Casting und Rolling.
No data wore found to bring output per man in casting and rolling up to date. olled output per manetric tons, and castetric
5- Timber Industry.
abor productivity in the timber industry was scheduled to increase byercent over0 level in logging and5 percent in manufacturing orubles6 rubles per man year,
In physical terms, output per.man in logging3 wasercent of0 level in spite of increased mechanization,as the result of the incorrect utilization of both men and equipment, gi/ In Frimorskiy Kray and Sakhalin Oblast, output per registered worker in terms of timber hauled was lessubic meter per dayThe average for the whole USSR may have been little higher. Judging from the blanket criticism referenced above, and other
In theears, output per worker per day ie scheduled to be increasedubic meters in the Far East,
* This relationship measures the total efficiency of thein contrast to the measurement of the productivity ofeach of the activitiesogging unit; that is,and
and presumably throughout the Achievement of this goal would probably approximate the goal originally setO.
6. Cotton Textiles Industry.
Under thc Fourth Five tear Plan,0 the productivity of workers In the textiles industry was scheduled to Increaseof0 level andercent At the same time, the productivity of equipment in various sectors of the textiles induutry was to increase by fromoercent, bringing output per unit slightly above prewar
It will be seen froo Tablehat. If the estimated rates of increase in labor productivity3 were achieved, the level of
Indexes of Output per Worker In tbe Cotton Textiles Industry
n^ Estimated between spinning and weuving
view of the data for thepprevlous year, thesebe
from the reported Increase9ercent in spinning andercent In
on the basis of prewar rates of Increase. 6g/
labor productivity waa atill below Uw0 goal. Present annual output per worker lo cotton textiles may bc as highO metric tons.* Inevel2 kilogram-numhers per autn per hour nay have been attained and in5 meters per man per hour, depending on the average density of yarn.**
During the Fourth Five Year, output per worker in the Soviet chemicals industry was scheduled to increase byorccnt over0 percent overfulfillment of the production goal in that period suggests that the productivity goal may have been attained, especially considered in conjunction with the annual rates of increase in productivity0 and subsequent years of lU,
Sulfuric acid is the only individual product for which both prewar and postwar data on output per worker could bc found. Thedata are5 and pertain tolants, where output man appeared to approximate6 level* tons perhe reported increase in average daily output of sulfuric acid per cubic meter of tower from kO kilogramsOilogramst9 should have contributedonsiderable increase in output per
labor productivity in the Soviet metal-fabricating industries is generally reported to be considerably above prewar levels, which were regained6 Although tbe degree by which thelevels are exceeded is almost unquestionably lower in terms of physical units than in terms of value, the complex nature of theof these industries makes comparison difficult because of the problems of measurement in physical Nevertheless, data
* Projected from prewar* Projected0 For example, thc Molotov construction machinery plant at Dnepropetrovskporcent increase in productivityi00 in value terms,percent increase in physical
shoving changes In World War II labor inputs into certain var If accurate. Indicate that the percentage increases incited in the same source vere onlymall extent tbe remilt of Increases in monetary
At any rate, significant advances in productivity in Soviet metal-fabricating industries should have resulted fron the priority given them In investment and the higher ratios of equipment to workers than in other industries, lhe increase in thc machine tool pool, vith the addition of more productive equipment is also cited by many sources. Somo indications are given of increases in labor productivity In individual plants resulting from new equipment and from Improved production
An an example of Soviet productivity claims In the field of metal fabricating, the index for the then Ministry of Machine andConstruction Is given in
Index of labor Productivity in Machine and Instrument Construction
d. Estimated froo previous year.
* Voznosenukiy's labor inputs for weapons vere compared with USJl/ The US-USSR ratio of inputs for small arms appeared reason-able, but Soviet artillery inputs vere about half those for the US.
Thc above oospores with the original goals0ercent0 for machine building in general,ercent for machine total
a.. Antifriction Bearings Industry.
Labor productivity in the Soviet antifriction bearings Industry may be estimated from that in State Bearings Plant Ho.oscow. As the largest producer, the plant is probably notof the Industry, but the industry average would be heavily weighted In its direction- Output per worker in Plant1ercent of output per workerjJ If7 level were equal0 productivity, as vas generally claimed forindustries, the index given in Tabicould result.
Index of Output per Worker in State Bearings Plant Ho.oscow a/
b7. j converted0 base.
If the increases in productivity13 paralleled those in3 wouldercentrnits per man-year. jj>/
Thus Soviet inputs were assumed to be for assembly time, or for only partially completed weapons. This may not rule out the accuracy of other data that Voznesenskiy used for labor inputs. Seeelov.
b. Automotive and Tractor Industry.
Under the Fourth Five Year Plan, labor productivity in tbe automotive and tractor Industry vas scheduled to increase0 level equalercent It would appear Likely that0 goal was attained, if not exceeded, from annual Increases Planned and apparently approximated ofercentndoercent Insofar as the level of labor productivity in the automotive Industry can be compared to that in the tank industry, the following estimates may be made.
L Iabor Inputs* tank1 werean-hours per An increase in tbe automotive and tractor industry to lU5 percent of this level wouldan-hours per ton This may be comparedurrent estimate derived from ejualogy to the USan-hour inputs per ton.*
Productivity in tbe automotive und tractor industryto increase byercent,0r
c asport Machine Building Industry.
Labor productivity in- buildlnctoercent0 Ihis ispercent greater than the roto of Increaseas
a whole, butn value terms, labor productivity inachine building industry was onlyercent of the averagether machine-building
The productivity goal0 was probably not achieved. Output per worker in one car-building plant9 vas reportedercent ofevel- In the some plant, man-hour inputsa gondola had been reducedy the end This may be compared roughlyurrentan-hours Inputxle freight car.** As cited by another source, man-hour Inputs into
* Estimate-fTrom7 equivalents, with Soviet productivity assumed asercent of US productivityan-year assumed
** Thesed for this estimate gives man-year inputs which were multiplied by anours per man-year.
locomotives similarly indicate an increase In labor productivity of aboutut the inputsoursare so high that they mist be regarded as
d. Oil Machinery Industry.
Labor productivity in oil machinery production reportedly more than doubled9 compared This would tend to indicatc^that productivity9 was atercent of thatnd if6 level were equal0 was moreercent of0 level. This rate of increase appears possible in tho light of increases in productivity in other branches of machine building and the trebling of the production of oilcompared
9* Railroad Transport.
The productivity of railroad operating personnel0 slightly exceeded the plan goal and, as indicated inalmostercent over0 Output peremployee is measured in terms of composite ton-kilometers; that is, of freight, passenger, and baggage movement. 9 the unit of measure was changed from operating ton-kilometers tohisepressing effect on the index.
The railroads in the territories Incorporated into the USSR in the west haveontinuing downward pull on the national productivity Index, but this might be changed by significant increases in traffic volute, Bince productivity appears to vary with traffic
* Using the same source which was used to derive thexle freight cars, steam locomotives would require0 man-hourc, and electric locomotives,0 man-hours if Soviet productivity wereercent of US productivity. ** Tableollows on An operating ton-kilometer Is in terms of distance actuallyariff ton-kilometer, tbe basis of freight charges, is based on the shortest routes possible, given existing track. The effect of this change in the USSR is given in. tfwt* urther discussion, see source
Output per Worker and Index of Output per Worker in Railroad Transport in the
Calculatedpercent Increase during the
Five Year Plan. 9j/ h. Interpolated0
c. Calculated on the basispercent increase/
di ower rateercent) than
Ohe index of labor productivity In Soviet construction which can be established from plan fulfillmenthown In*ontinued failure to attain plan goals in spite/ lhe Fourth Five Year Plan goal0ercent0 productivityercents compared with the level achieved0ercent4 The increase planned50 was/ Con- idering the achievementshe level attained5 will not be much more.
* Tableollows on
Index of Output per Worker in Construction In the
on the basis of anincreaseQj/
he output per man-year In constructionwork uas0 rubles. 3 it had risen to an estimated ftl.ftOO rubles.*
* One/0 nan-days of labor in basic work per million rubles worth of construction and installation work,9 rubles per man-day. ays per year, calculated from1 Plan, this0 rubles per roan-year. This figure was projected3 by using the index in Table 1ft. These figures are probably5 rubles, which were being introduced for use in construction estimates at this time. 'This view Is reinforced by tbe fact that the man-day ruble output calculated from it0 dataubles)2 percent, whereas the Index in Table 1ftt73 percent oft0 level t0 data were calculated fromtl Plan).
Since labor productivityatio derived by dividingunits by employment units, tbe significance of the ratio is affected:
any unreliability of the production data or crudity inprocedure adopted for aggregating individual productscombined indexes of product ion; that is, in aggregating theof various chemicalsepresentative chemical index.
any unreliability of the employment data orof the employment series for the purpose intended.
2. Measures of Production, a. Prevar.
It is generally agreed that Soviet published basic data on physical volume of production7 vera reliable and published in sufficient detail to constitute the basis of significant measures. Soviet value indexes, however,uble prices tototal value,rossly exaggerated picture of production increases because they gave inflated values, because new products were artificially "priced in" by procedures which exaggerated their effect, and because they duplicated the value of products which are re-used in/ This latter problem arisesasic product is used ininished product, such as when pig Iron isinto steel end items. ross value index would include the value both of the iron and of the fabricated steel, thus duplicating the value of the pig iron.
The difference of prewar productivity indexes based on value from those based on physical volume is shown inhf
* Tableollows on
Indexes of Production in Selected Industries in the USSR
An additional estimate of the difference between volume indexes and value indexes is furnished by/ by comparing the Soviet official gross value index with an estimated physical volume index. Tola shows an advance in the official (value) index47 ofercent as against onlyercent in Hodgman's calculated index.
. Because of these difficulties which underlie the Soviet official index of .productiono attempt is made in thisto"analyze trends The major analysis centers on the. The indexes are based on the0 in order to tie"the current period to the level attained Just prior to World War II. No attempt is made to trace the trends in the abnormal war years0
Postwar Soviet official figures relating to the trend in production are probably less'exaggerated than prewar series, but there is somo controversy as to the extent to which the methodologicalhave been eliminated, lhe indications of improvement are:
fter widespread criticism of the official series based on gross valueubles, new pricing procedures were recommended. These were based on new current price schedules, and
the inflation which had been Introduced by the previous system of pricing in new products was minimized. It seems probable, however, that the present system of weighting comooditics is not strictly bused on current prices- Otherwise, the price reductions of recent years would probably haveepressing effect on the productivity
A second effort at improvement has been the tendency tothe gross value principle in favor of some weighting system based oo the net value added. Industry manuals available to/ which contain quite specific instructions as to methods to be employed in the statistical report log of production, place emphasis on the net value principle.
It should be emphasized that the Soviet production andclaims, which are the backbone of this report, are from published figures and that information ie lacking as to how thc production and productivity reports required by the manuals mentioned above areto produce the official published index. In the absence of such knowledge, about all that can be said is that Soviet production figures on which productivity estimates are based provide the only means of attempting to fix the general order of magnitude of theof the Soviet economy.
Since the absence of other data has led all analysts oftrends ln the USSR to base their conclusions on Sovietfigures, the.principal differences in-the measures computed arise not from^differences-in facts as to the volume of production of individual conmodlties but from methodological differences in the statistical procedures employed to build up aggregate indexes. In order to provide some independent check on the Soviet official index, productivity has been calculatedther production indexes, using the some employment series inndexes. This comparison, shown in Tablerings out the differences between the Soviet official index,ndex, and the CIA index. The differences shown arise mainly from differences in coverage and differences in statistical method. Ihe two non-Soviet Indexes are based insofar as possible on physical production statistics combined insofar aswith estimated value-added weights. The number of productsis different, and it la apparent that both independent indexes are less comprehensive than the Soviet Official index, which is probably based on fairly complete coverage. This comparison is presented not in order to rate one measure or the other measure as superior but to point out the divergence which can arise from different uses of the same data
and to indicate the general agreement In all three as to the cxisteronounced upward trend even though the steepness of the three curves varies-
oes net show the differences ir. tiieexes from which the productivity Indexes are derived. Thisis Shown in
Indexes of industrial Production in the USSR Selected
It will be observed that8 there has been substantial agreement between thc CIA index and the official index, the former increasingercent and theercent8
3- Measurement of Employment.
lhe employment component of the productivity index is not subject to the same statistical difficulties as the production It is highly probable that the employment figures used represent production workers only, the definition being very similar to that of production workers ir. US statistical usage. For some purposes it would be revealing toomparative index based on total employment, thus including the bureaucracy and nonproducers. The tendency to overstaffnproductivc Jobs has at various times led to severe criticism of the system of manpower utilization, and for thiseasure which would reflect the fluctations in this nonproductive group would give some clue as to waste of manpower. Hon-
productive employment cannot, hovever, be derived from available material, and information On this subject is therefore of an indirect and qualitative nature.
A second unsolved question relates to the type of average vhlch is used for reporting employment. This, however, should not affect the trend of an index unless the definition vere changed during the period covered.
A more serious statistical flaw in the productivity measures probably arises from the system of reporting the inputs of slave and prisoner-or-var labor. It appears from the behavior of some of the industry indexes during periods when prisoners of war were beingthat the output of this group vas included in the production series but that their labor input was excluded.
For the purpose of indexingarticularly appropriate method is to disregard value and to vcight indexes of physical volume by the size of the labor inputs used. Otis would take the following form for an individual industry:
Index of Labor Productivityutput per man in given year (l)
p Output per man In base
Physical Volume of Production
Physical Volume of
The aggregate for industriesn would then take the form:
Aggregate'lp E'^B" *
Such an index has technical advantagesroductivity measure, the principal one being that it is unaffected by changes in the price level. It may be interpreted as follows (to use the simplest case when the number of vorkers is constant); If the output per workerorkers in industry I' Increasesercent of the base year and the output per workerorkers in industry I"
increasesercent, then thc aggregate product viiy o! thuIncreases as follows: l|4
Aa Index of this type was calculated by the/ except that production may have been ex-
pressed in value terms. There is, however, no evidence that it was published. Its abandonment was reccccrcnded in loJtS, at which time the stated reason was that it did not yield results sufficiently different from the older method to warrant its use- it would appear, however, that an additional reason was that it did notufficiently optimistic picture of the increase in produc'-ivity.
Regardless of whether the Central Statistical Office still uses this principle of aggregation, some of thc manuals examined recommend it for the aggregation of products in railtiproduct plants, and the planners may base their reasoning as to productivity on such-
(JAPS IH II'.TEIXiaENCE
Information in lacking aa to the components oi" thc Soviet general index of industrial productivity and as Co its method of compilation. It is not known whether thc Central Statistical Office divides theproduction index by un eaployment index or whether Itroductivity Index directly from plant and ministry reports. It is believed, but not certain, thai coverage is industry-wide and not selective-
There is no way of measuring the extent to which production figures are inflated by the inclusion of unfinished or defective material, but scattered evidence indicates that thisactor. An index would not, of course, be affected by such inclusion if the proportionconstant. There is reason-to believe, however, that the amount Of waste in industry in the USSK has been reduced.
The extent of the inclusion of the production of prisoners of war and slave labor is not certain.
Although it is fairly clear that Soviet productivity calculations are based on production workers only, no statistical Information is available on nonproduetlon employees. An index based on the total production and nonproduetlon employees would be Influenced by changes In over-all efficiency and would give valuable information on bureaucratic waste of manpower.
Study should be devoted to the managerial contributions to efficiency.
Specific information is lacking on productivity in food processing, and information is spotty on chemicals. Facts as to productivity in water and road transport are not sufficient to construct an index. Information is lacking on some sectors of metal fabrication, electric power, and conmun-ostions.
The extent to which the productivity figures announced forgroups cover all the productsinistry or only theproducts Is not known-