Created: 11/12/1954

OCR scan of the original document, errors are possible





CENTRAL INTELLIGENCE AGENCY Office of Research and Reports



Io April andeetings uere held ln Moscow to discuss tbc Soviet railroad transportation plans* and the following year6 and to review the accomplishments and fallings Two Important speeches were made by Kagaaovich, First Deputy Chairman of the Council of Ministers of the USSR, and two by Boehchcv, Minister of Transportation. Study of these speeches, and of others by subordinate railroad officials, has given very useful material for the computation of past performance aod for ao assessment of the likelihood that future plans will be fulfilled. Tho speeches give little basic statistical material directly, but by comparing one speech with another and by mathematical analysis It has been possible to build upevealing picture3 railand to gather useful data on plans for the.

It Is believed that the figures contained ln tbe speeches ore reliable and were not released for the benefit of Westernorganizations. The facts given are those which railroad workers need to know to advance the government's plans, and their disclosure seems motivated by the necessity to denouncefeatures3 operations in order to goad workers to greater efforte. Furthermore, certain facts can be checked by independent Western observation of Soviet rail operations. Other Tacts are consistent with information in Russian technical books issued for the instruction of Soviet railroad men. Finally, tbe figures derived from the speeches seem internally consistent, and it bos been possible to obtain certain key figures ln two or more independent ways.

This memorandum by no means exhausts tbe basic material Much additional information can eventually bc derived from these speeches and from fragmentary older data, including actual traffic oo some key individual Soviet lines, but the research required ls laborious and will notull study to be

completed for some tine. In its preliminary form, however, analysis of the speeches has already filled several important intelligence

gaps. The memorandum Is therefore being issued at this time so that thc methodology and results obtained to date may be commented

on by others and the conclusions employed in evaluating Soviet

attainments, capabilities, and intentions.

- iii -


CIA/RR rtf-4o4



3 the Soviet railroads vere operating under strain and close to present capacity, with an Intensity of utilization per mile of track and per car about three tines tbe US level, nevertheless, they handledercent of total Soviet traffic, andillion metricf goods. With an average haul of 7U6 kilometers, thisotal work performance ofillion ton-kilometers. Both tons originated and ton-kilometers vere the highest in Soviet history and vereercent above previous CIA estimates.

Study of the speeches of Kaganovich and Be shehows that record movements of bulk consnodities produced at least moderate shortages of gondola and hopper carshile complaints were also made of inadequate supplies of refrigerator cars and cattlecars as well as shortages of containers to move less than carload quantities of consumer goods freight* Locomotive supply was adequate, but it was evident to the Soviet authorities that more powerful types must be built to haul the heavier trains now being run, and prototypes of such locomotives were constructed It seems clearolicy of forcing higher performance out of locomotives waspushed during the year and thatolicy did not have the approval of some leading technical men, who feared eventualof equipment.

The year showed some retrenchment In plans for construction of new lines, but nevertheless there were significant additions to the rail net in the South Siberia-Turkestan area. Emphasis vas put on

* The estimates and conclusions contained in this report represent the best judgment of the responsible analyst as* Tonnages are given in metric tons throughout this memorandum.

Improving present trackage in key areas, with the laying of better ballast aod heavier rails, so that train speeds might be Increased lo the future. Moderate additions were made to the electrified line, which le slated for further expansion.

The campaign for haulage of heavier freight trains vas strongly pushed, and it Is calculated thathe average 8oviet freight train hauledons of goods, orittle less then did the average US train* This figureotable increase over the freight tonnagesew years ago. Turnaround time vft3 reduced only moderately during the year ln spitetrong campaign for Its reduction, but nearly attained the planned figure.

The total intensity of utilization of the railroads in the USSR was for greater per mile of track and per freight car ln operation than was the case in the US. Much less spare capacity was therefore available for handling any emergency movement of goods. Average train epeed, although good compared vith previous Soviet performance was not up to US standards, being hindered by poorer tracks, less traffic-control equipment, poor brakes on the older cars, and smaller locomotives.

Considerable progress vas Bade in turning tbe active car fleetxle ooe, with freight cars comparable to those of the US. Plans were ln force foraxle cars as they wore out and for making new constructionxle cars of an advanced type, ofxle cars from main routes can at some future time greatly Increase traffic efficiency.

The following tabulation gives the more important flgurea derived from thc speeches3 Soviet railroad performance:

L. Ton-kilometers: illion, which is the highest total in Soviet railroad history. (The previous CIA estimateillion ton-kllometero.)

2. Tons originated: illion, which Isecord figure. Of this, coal wasercent.

3- Average length of haul: 7k8 kilometers. (The previous CIA estimateilometers.)

4. Average length of freight car movement from one loading to the next: ilometers. This distance Includes empty haul.

5- Average turnaround time for freight cars: 70ecreaseours2 andours under

ut about the same as (The previous CIA estimateays No estimate had been made)

6. Breakdown of turnaround time: ercent at stations en5 percent at distributing8 percent at classification3 percent at stations where loading or unloading work is carried out,4 percent actual time in motion.

J* cal speed (speed of trains between3 kilometers per hour.

ncluding stops en route):

ilometers per hour.

daUy carlD0dtne: arsxleigure isercent)

ut^Lmlleafie of Russian-gauge lines,and the like: ilometers. This mileageunder construction. (The previous CIA estimate was

11. Average freight density: illion tons of freight over the average kilometer of line per year. This0 tons per

i f ?ULaWdtraioloadons, this 7 tfith ereatest traffic

vf^.CC^tedb0utrainB Der ^directions, and double track) past any given point,ons of freight per average train on such routes, These figures seem to agree with actual observation.

r^cooCBFSXlC unlts> in actual use at6

t0taJ freight car park (in use, laid up for repairs, t hysical units, made up ofxle unitsxle units. xle equivalents this works outotal parkars. These figures rest on the assumption that there is little laid-up reservexle freight cars, considering Soviet traffic needs, but thattla^ee reservexle cars. (The previous CIA estimatehysical units.)

f "tual carrying capacityxle units2 percent. On the basis of actual Western trafficplus computation from data in the speeches, aboutercent of the cars running seem toxle and onlyxle. The carrying capacity ofxle carittle over twice as great as that ofxle car, and therefore tn essentials Soviet railroads are now almostxle basis.

15. Division3 traffic (ton-kilometers) between various media or transport: rail trafficercent of total;ercent;ercent; other (truck, pipeline,ercent.

I. Soviet Plans.

Very considerable increases in carloadings, in tone originated, and in ton-kilometers took placebut the increases wereecreasing rate from each year to therojection of tbe rate trend tends tolanned increase in carloodings ofercent3 as the Soviet goalj which has been derived from Beshchev's speeches.


Plans were madeecrease of ao unspecified amount in the average length of haulk aodharp upsurge ln the number of heavy trains. Delivery of new-type equipment consisting of heavier gondola cars and more powerful locomotives was also to begin this year.. Further Improvement of track and of automatic signalling equipment was to be made, and tbe electrification of key lines pushed. Train speeds were to bo increased and weight norms raised.

Goals for later years were much more ambitious. percent Increase in the level of freight handling0 wasby Kaganovich, who also spoke of deliveries of new and more powerful locomotives ineriod amountingteamlectrics,lesels, plus large unspecified numbers of gondola cars (uponsnd refrigerator cars.

The rationalization of hauling by the elimination of cross-hauls and thc building of factories and flour mills in regions which produce raw material and which also use the finished product in order to decrease the need for moving as much material as possible,were also stressed. Heavier trains, faster schedules, better roadbeds, and better traffic-control devices, as weLleduction in time spent in yards and terminals, seem to be the contemplated answers tl the need forperCcnt increase in tons to be carried.

- I* -

II. Probable Performancek and Later Years.

Analysis Bade of Soviet traffic figures released for the first half4 indicates that railway traffic* will probably beercent3 in tons originated and in ton-kilcmeters. It is estimated5 should shew about the.sane growth. Thereafter, growth may beigher rate. Although there is an ample number of tank andhortage of heavy gondola, hopper, and refrigerator cars presently limits traffic increases in such key items as coal, ore, building materials, timber, andfood products, aod the program for expanding output of such cars above present levels is not likely to show major results More powerful locomotives are also needed, although thereood supply of those of ordinary type. Prototypes of the new locomotives were delivered3 and the first partut quantity production is probably many months in the future.


Plans were madeonsiderable increase in the number of heavy trains It is reported that, for the first half year,ercent of the trains were heavy cues. The significance of this is not yet clear, however, because the total number of trains ia comparison with the corresponding period3 Is not given, the norma for train weight are not generally available, and the degree to which some trains night have been underloaded to furnish tonnage for heavy trains and thus allow paper compliance with the plan is not yet (mown* Nevertheless, it ls likely that average train weight is still increasing, although nowmaller rate than Sovietwould have the world believe.

4 goals for the reduction or turnaround time and for the increase lo train speed are not being met and will probably not be met,ittle progress will be made. No appreciablein average length of haul seems likely5

For the longer term, the USSR desires an increase in traffic volume ofoercent as compared This goal will probably be met although at the cost of much effort..

Attention is called to the fact that much electrification of main line track is scheduled for the area from the Volga to Western Siberia ineriod. This will inoreose traffic capacity, but also vulnerability, thus giving peacetime strength and wartime weakness.

III. Derivation oftatiBttcB.

The baste procedure for deriving3 statistics was to find in the key speeches those figures and relationships that could be used in obtaining such statistics, and to cross check conclusions for consistency wherever possible, ln order to minimize errors in interpreting the data furnished or ln the manipulation thereof. Cross comparison has also been made with available Soviettexts, recent articles to Qudok (tbe Soviet railroadnd information in the bands of commodity specialists.

1. Ton-Kilometers.

Kaganovich ln bispril speech says, "Ve have railroad routeshich tbe freight intensity reaches fromoillion tons per kilometer with an averageillion tons." The context makco it plain that the average referred to Is that for the USSR and not for some particular route and that it Is But hestates that the station-to-station length of line for the USSRilometers. Since Soviet statistical texts show that average freight Intensity is obtained by dividing total ton-klloaetcrs by total kilomctrage of line (station tootal ton-kilometers con be obtained by multiplying kilometers of line by average freight intensity. This procedureillion ton-kllcotetersAgreementrior CIA estimateillion ton-kilometers Is good, thc derived total beingercent larger. Part of this difference Is attributableeported Increase ofercent in average Soviet length of haulhen it might have been expectedlight decrease would have taken place. In spite of the useounded figures io deriving the total, the result is believed to be accurate toercent. reater margin of error would probably be reflected in differences lo tons originated figures derived by different methods, but, as shown below, the differences are minor.

2. Tons Originated.

The Kaganovich speech ofpril atu'cs, ln summary, as follows: "There are at any one5 million tons of goods on the railroads. The value of these goods lsillloo rubles. There is an average delay inon, as compared witn schedule,6 days. Eliminating this delay wouldillion rubles tied up in the value of goods in transit." Tt follows that we day

from these figures the average tine for goods Id transit, the formula being 26 6 day. The result is an averagekk


days in transit.

This time (which excludes loading and unloading, empty haul of car, and time io technical stations for the emptyivided into5 million tons of goods in transit,ons originated per day,ons for the

A cross checkhis figure is possible, since the 8U0 billion too-kilometers already derived can be divided byiven by Kaganovich as the average length of haul. Thisigureons, or less than two-tenthsercent more than by the first method. The probabilities are that the firat method is slightly more accurate, since fewer rounded figures are used in its derivation.

A third general check as to order of magnitude is possible from Beshchev's speech, in which he states that increasing the average load per carercent would yield an increase of overillion tons more hauled per year. That is, he implies that the tons originated figure isillion tons.

A fourth check of considerable validity has alsoby using Kaganovich'spril remark thatof the freight carried was coal. Twenty-eight percenttonsons. Consultation withspecialists revealed that theyigure forroductionalenkov speech. The specialistsfor coal used at theor loaded directlyhipsent on thespecialists made u

corresponding subtraction fromons moved, for Polish and Chinese coal imported into the USSR and hauled on tne railroads but of course not countedart of Soviet coal production. Further reduction was made for coal possibly double-counted inthrough use of rail-water-rail hauls or through picking up for on-carriage of coal previously hauledtockpile- The figures were then compared. The figure derived by subtracting ninerrcm reported production wasercent less than the roughly corrected rail transport total. In view Df the tentative character of the corrections made, the agreement is excellent. Unless very consistent falsehood is being uttered by

various Soviet officials at various times, these general cross checks serve to validate each other and to give credence to the figures derived.

3- Average Length of Haul.

The Kaganovlch speech ofpril specifically says that average length of haul3ilometers. The figure appeared io connectionengthy denunciation of cross-hauling and unnecessary movements of rav mater lals to factoriesom regions where the finished products arc consumed. Examples are also given of major commodities for which average'length of haul has increased notably la recent years. Data given are con-

already koovl'Soviet statistics for

is thought

th.ofpotation does not definitely show whether

clLlofTs Sa?S2 tTntatlvHoo!

t.are meant. This conclusion would be consistent with quotations for average speed given in the same and other speeches, which clearly must be operating figures on the aclual tracka used. Irrespective Of the ehortest theoretical distance.

k' Average Length of Freight Car Movement from One to the

*Yldcrit frOB "Wre of Soviet rail shiprents that

much hauling of empty cars must take place, even though there Is

* VCduCe U- w* necessJY^run^tJ a

EJ/Eit ilfield or refinery

aaodgravel, and similar bulk commodities also often Involve empty hauls, as do move-

" frigerator and cattle car.. Seasonal traffic1 generate empty movements,udden demand for cars to move grain,

;arcomeegion than

JJ if ommodities which at that time require transport to the place of grain production. In general, then, an unbalanced flow of 8 movempnt of equipment adapted only for carriage of special corneal ties, causes empty car movements and resultseed for more cars than would otherwise be the case. The ratio of empty-car movement to loaded movement is there Tore of importance.

- 8

For thc USSR this figure can be derived directly, and checked indirectly. Aa Is Bhovn in the discussion on turnaround tine, Kaganovich gives data whichreight car turnaround tineours3 and fumishea information froa wnlch it may he calculated4 percent ofours was thear was actually in motion. Actual running time, therefore, must have beens will subsequently appear, however, average technical speed (speed between stations and excluding time for stops) has been found to be of the orderilometers per hour. It follows that the average car ranilometers from one Loading to the next, and tnat since the average loaded movementilometers, the average empty movement wasilometers. The average empty movement is4 percent of the average loaded movement.

Pavlov, in Gudok for ILtates that "empty car movement is nowercent of loaded This is certainly consonantomputed figure4 perceut

Beshchev, in his speech reportedtates that "out of everyars oo return rune, therempties. This percentage has not changed for fifteen years." This remark tells nothing directly about average length of freight ear movement but Ls nevertheless usefultarting point for statistical approximation.

First, it Is to be notedarge part of the empty movement must be return of coal, oil, timber, and similar cars from point of delivery of merchandiseine, refinery, or mill to pick up another load. That is, the return run ought to approximate

the outward run ln kilometers. Interchange of cars at unloading points does not change the ratio of empty run to loaded, as long

as tbe commodity flow pattern Is stable. Empty runs of boxcars, however, should be decidedly shorter than loaded runs, sincecarry varied freight in triangular movements, and run to the nearest point of demand as different cceaiodity movements dictate, rather than operatehuttle serviceew fixed areas of productionelatively few main consuming points. Offsetting this, on the other hand, runs of oil, timber, and some other bulk freights are much longer than the average. In summation, it seems probable that shorter empty runs for boxcars and longer empty runs for tank cars, gondolas, hoppers, and otner special cars will tend to offset each other and that an empty haul will be similar in length to thc average loaded one.

* Footnote references in arable numerals are to sources listed in the Appendix.

- 9

With this assumption in mind, on approximation of average length of freight car movement from one loading to the next can now be derived from the Beshchev data. For everyars moving loadedlll be reloaded there without empty run. These vlll then have moved the average loaded distance offrom one loading to the next. ars) will have to move aoilometers each as empties to another station before being loaded again, making the average distance for theseilometers from loading to loading. For theotal loaded movementilometers and an empty movementl8 kilometers results,ombined figure8 kilometers. Thisilometers per car. Agreement vitherived from calculations based on the Kaganovich speech Is excellent.

5- Average Freight Car Turnaround

Turnaround time is the total period that elapsesar to the next loading, including time spentand unloading, time In classification yards, and timeKaganovich, in his speech oflacespeeding up the turnaround time for freight cars asof handling the increased tonnages scheduled to move onUrates conditional

oadlQe and unloadingr thatlowdown of movement of such cars en route, and demands that great improvement be made. He states that "the ^rnoroundar can be speeded up through the utilization of all the above-mentioned reserves (meaning points where time can bey aboutours orercent, and not byercent as proposed by the Ministry of Railways." He recognizes, however that thisifficult task and one which will take "several' years" for realization.

OUrB iSerccn*3 turnaround time, thentime must haveours,? days, andplan to reduce it eventually to 5

Cross check is possible, for Kaganovich states in an earlier

PZ* thath Car SPcndsours idling at

Ire cwiiJS rnaround Technical stations are classification yards and distributing railheads.. But Kaganovich also gives at anotherercentage breakdown of the total


turnaround, time, from vhlch we learn that the average freight car spends Its time as follows: ercent at stations en5 percent at distributing8 percent at classification yards,3 percent at stations where loading or unloading work is carried out. By difference, actual time in motion must4 percent of the total.

It follows that if7 hours at technical stations3ercent5 percent) of turnaround- time, then the Utter4 hours, which is practically Identical withours previously obtained..

Further indirect check ia possible through the computations for technical speed that will be found In the following paragraphs. Comparisons with data for earlier years found In Soviet technical works, and reported improvements in turnaround time from these earlier years3 mentioned in variousok articles, also3 turnaround time 'almost Identical with that above. From such comparisons and articles, CIA had estimated prior to receipt of the Kaganovlch speechurnaround timeays would be likely to be attained in the.

Beshchev, In his speech reported forIndicates that turnaround time3 wasours less thaond aboutours less thanelease4 from the USSR Central Statistical Administration covering theonthstates that turnaround time was reduced slightly during the period, but that it did not meet thehisood indication that theercent reduction called for by Kaganovich may be hard to attain, and that theercent reduction called for by the railroad experts may be more realistic.

6- Average Technical Speed and Average Commercial Speed.

In his speech ofaganovlch deplores the fact that in past years technical Bpeed (average train speed excluding stops) has not risen, and he blames this on over-generous norms established by the railroad lines and the Ministry ofe states that these authorities "when determining the time the train will run between stations, have allocated unnecessary margins in calculating running times.0 He states thathese marginsercent and that their elimination will result in raising technical speedilometer per hour. Then, by simple proportion,


technical speed shouldilometers per hour, but this figureittle low, and it must he home in Bindlight rounding intatistics canariation of as much asercent in the derived technical speed. Dr. Hollaed Hunter of Haverford College, who has made extensive study of Soviet rail transport, calculates that the technical speed08 kilometers per hour. 2/

Having In mind Kaganovich's statement that during recent years technical speed has notossible upper limit of aboutllometerB per hour might be set. Data on hand do notefinite choicekilometer speedkllomoter speed, since one cross check agrceB best with the slower speedecond with the higher figure...

Io the chapter dealing with average length of freight car movement, distance travelled from one loading to thc next is derived by usingilometer-pcr-hour speed. This is comparedirect figureoviet transportation expert andigure indirectly obtained by an approximation method, fron certain data given by Beshchev. All agree very closely, but the agreement is destroyed Iflloneter-per-hour figure is increased moreew percent.

On the other hand, commercial speed (that is, speed including time spent at stations ens given by Kaganovich5 kilometers per hour less than technical speed. On the basisechnical speedilometers per hour, commercial speed should beilometers per hour.. Agreement with calculation is only approximate, however, and cannot be considered very satisfactory. The time spent at stations en route is given by Kaganovichf total time and the time actually spent in movement (derivedesidual byU percent of total time,ombined figureercent of total time as applicable tospeed. Multiplyfngour turnaround time total by thisombined figureours results. This figure dividedilometers, which has already been calculated as the approximate average length of freight car movement, ought to give commercial speed. This calculation, however, makes suchilometers per hour insteadilometers. The difference isercent, but it is hard to account for. Kaganovich can hardly be in serious error in his statement that the difference In technical and commercial speeds5 kilometers per hour. But failing such

- I? -

error, tbe figureechnical speed of aboutilometers an hour in order5 truly represent tbe difference between technical and commercial speeds if the relative percentages shown, for time ln motion and time ih stations are accurate. igher average speed for an Identical time, in movementreater distance moved. Since loaded distance is fixed (being given byhe increase must be In empty haul. echnical speed ofilometers per hour, the "total distance traveled works outilometers, while loaded haul is given directly by Kaganovlch as jk& kilowters, making emptyilometers, orercent of loaded haul. This directly contradicts percentages of empty to loaded haul derived frca Pavlov and Be she he v. and gives total length of unproductive' freight car movement of such magnitude that it certainly would have called for major denunciation In the speeches, if real. No such denunciation took place.

Tentatively, then, it may be assumed that there Is some rounding error .In percentages given or derived, and that true technical speed ieittle greater thanilometers computed, but decidedly less thanilometers. Calculation might also have been thrown off by seme difference between Soviet definition of technical and commercial speeds as given in Russian texts and methods of compiling figures actually used by the Ministry of Pending receipt of further data, the discrepancies cannot be resolved, and the figureilometers per hour will be kept for technical speed for lacketter.

7- Average Dally Car loading.

In theay speech, Kaganovlch discusses the underloading of cars and states thatecrease of loadingon per carransportation loss of more thanillion tonshis means, of course, that more thanillion cars are loaded per year. Soviet statistics on carloadings are kept Inxle units Irrespective of the real size of the care actually handled. Therefore, it can be calculated that morexle units are loaded per day. The actual figure is probably slightly higher but is unlikely to benits, since otherwise the average load per carittle low,ounded figure for annual loadings vould not beillion butillion cars. ough cbeck Is possible by use of Beshchev's statementilometer increase in average haul3 decreased dally carloadingsars. Then, bys0 hours turnaround time is to hours



Jet, which Hakes the tine5 hours. Calculation of tlae taken

ilosicter extra leaded haul, plus pro rota allowance for apty haul and for added tlae ln technical stations,our mb fron the greater' distance. The agreement shownrloadlng figure cannot be far ln error.

8. Statlon-to-Statlon Mileage.

Kaganovlch, in theay speech, says "our railway conveyor is spreadilometers." Context makes lt virtually frtain lhat ho Is talking of total statlon-to-statlon distance, nioflng number of tracks and ignoring sidings. The figure- is lentical with

9- Average Freight Density.

In Kagaoovlch'spril speech, he says "we have railway xitea on which the freight density reaches fromoillionper kilometer with an averageillion tons." Context ikes it clear that the average refers to all lineseight density, as here used, means the combined total tons of eightoint in both directionsear and showswhat percentage line utilization bears to line ipaclty.

A yearly averageillion tons ia equal0 ma of freight per day. Thc average train carried3 about

ons of freight, thus an average4 freight trains must .veiven point Inour a. ThU would be roughly 0

eight tratnn tn each direction per day.

The lines or maximum freight density are known from Soviet alements to be lines ln the Donets area and the western port ofTrans-Siberian line. For the latter, train counts exist, from .lch it Is possible to Infer average train loads ofof freight.- Using this figure,requency of approxl-.telyrains (both directions combined) per day alio obtained omreight density0 tons would be rived per year. Agreement with Kaganovich's figure oflllloiIs excellent.


10. Number and Ctrniposition of Cars per Train ana Average Gross ond Met Weightrain.

Computations In this field rest partly on direct observations of Soviet trains,artly on data ln Soviet indications, and partly on matter Lc tbe speeches under analysis. Kaganovich says in hispril speech that "Wc must have heavy trains and these need powerful engineshese engines oust pull trainsons insteadons as at present."

Three interpretations seem possible ln regard to the last part of the remark. Kaganovich could meanresent "heavy" trainons, but this is contrary to much evidence that trainsr more tons often move on aomc coal lines and ls aleo much below "norms" for freight carriage prescribed for trains on various lines as published in Qudok, He could mean that present engines are unable to pull trains weighing moreons, but this is contrary to numerous reports in Gudok of single engines hauling much more than this and also disagrees with "norms" seen for carriage on certain lines. Finally, he could mean that the average train nowons exclusive of engine and tender (which would not be counted in Soviet practice) This lastseems plausible and will be used hereasis for calculation.

Train counts byin the

Moscow region and on field trips nave been roughly tabulated,to show that an average freight train Is made up ofcars,oercent of whichxle ondnits. U/ It seems rrnuonable to assume that trainslines not seenmay average somewhat

smaller, so for this computation an average train has been taken asars. It will rirst be assured thatrain hasxle cars, and that tt Is accordingly composed ofxle andxle units. Empty car travel is about one-third that of loaded travel In kilometers. On the average train, three-fourths of the cars would then be loaded and one-fourth would be empties movingoading station. We then haveoaded large carsoaded small ones, equal toxle units per train. But on average load ofons perxle unit is computable from tlic Kaganovichxle units loaded per yearillion tons carried). The average

llO-oax trala would thenoudons of goods. If the ratioxlexle cars Ls onlyonstead ofhe average load would still work out toons of goods by analogous computation.

Assuming thatercent of the carsxle,axle cars with an average empty weight of5 tons each, andxle cars each averagingons willombined car weight per trainons. Subtracting this fromco train weight figure, wereight loadons which is onlyons less than that derived from average ratio, weoadons, orons more than that derived from load factors. The close agreement makes it seem likely that Kaganovich was in fact talking about the gross weight of tbe average train, and an estimateons of freight per train seems therefore to be fairly reasonable, even though thisarked advance over the probable average weightsew years ago. Incidentally, the computationsxlexle car ratio of abouto3 seem plausible for the USSR, since this ratio gives minimum difference between the load derived from average weight per car sod the load derived byempty car weights frcsion train weight given by Kaganovich.

Some general considerations are also pertinent to this train weight problem. First, it seems clear from the speeches and from Qudok items of the last year or two, that Increasing the number of carsrain and the load per carthat Ls, the total freight tonnage per trainasajor goal of the Soviet government for eame time.. It la unlikely that no progress should have been madeield marked out for special attention. Progress up3 wss undoubtedly less than planned, however, for much of the furor over "above-norm trainB" must have been merely paper Kaganovich implies as much. It was onlyhe USSR apparently realized that it had been ordering that heavier trains be run while simultaneously paying yard men by the number of trains sent out, rather than by their weight. Obviously,eavy trains ratheright ones meant lesu pay for the yard men, and paper compliance with an orderertain number of heavy trains merely meant that some trains were made heavier than norm, while others were correspondingly underloaded. The change lo method of remunerationave hadresult on the weight of trains dispatched last year.


Another factor bearing on the operation of heavy traino has been the grovth in the numberxle cars. Inadequate siding lengths and restricted yard truckage are limitations on the site of trains that can be handled on many Soviet lines and are factors ccmiwnted on in some recent minor speeches. xle cars take up aboutercent more track length thanxle car of equivalent carrying capacity, according to Soviet technical works. It follows that as larger cars take the place of tbe small ones, trains can carry more vlthout lengthening yards and sidingand it ls known that the percentagexle cars has grown.

Iteasonable inference froa the Kaganovich speech that the USSR now loads almost as much freight into ao average train as does tbe US.* The comparisonittle misleading, however, since US railroads handle much greater percentages of light-loading manufactures and leoa-than-carload freight, move much produce by refrigerator'car, and also Indulgear greaterof hauling of empty cars, in line vith the different US Idea of true economic efficiency. Since Kaganovich is nowreat increase in Soviet consumer goods traffic, including increased refrigerator car movement, tbis may ln the future cancel out to some degree the emphasis being placed on heavy trains. But Soviet railroads are at the present time still essentially haulers of bulk freight and may therefore surpass average US train loadings in the near future. This vill not mean that they have surpassed US lines in efficiency, but that theyore primitive economy.

Before leaving the subject of trains, it may be well to point out that tbe constant barrage of propaganda ln the Soviet press about so many thousands of heavy trains dispatched and about surpassing the "norm" for freight carriage ls often quite misleading. In the first place, "norma" differ for each line, and probably for various types of bulk commodities moved, andnorm" therefore means almost nothing concrete, especially since they can be raised or lowered without public notice at will. Where vehe Information ls useful, and data have been collected from Russian newspapers and broadcasts indicating some apeciflo ones, varyingons of freight per trainine io the Moscow areauspected normons of freight per train for coal trainshe Pechora railroad and one ofons for certain Baltic area lines. Beshchev said4 that weight norms have been substantially increased

* The US averageetric tons of freight per train


Furthermore, there is strong evidence that much of the "above-norm" transport bas hitherto been accompalnedrop in tonnage lifted by other trains on the same Lines, sotatement thatercent of the trains were "above-norm" on some partlcuLar railway does not mean that on the average the norm for haulage per train was exceeded. The general context of an article may convey an impression that the reported attainment waa real or that it was merely traffic Juggling to evade penalties for noncQmpliance, but in most cases the articles do not give enough facte toomplete and honest picture. The statistics cited may be the truththey usually arebut they are not the whole truth.

Pravdaase in The statioomaster of the Sverdlovsk Routing Yards is quoted as saying that in his yard heavy-load trainsone of freight above the plan during the first quarterb:out underloaded trains failed to meet their normsons, and thatesult the station failed to fulfill its plan for the quarter.

Number of Cars ln Actual Use.

By Soviet statistical definition, the number of cars ln actual use at any given moment Is derived by multiplying average carloadings per day by average turnaround time in days. Average carloadings3 werexle units) and atime7 days has been found. It follows thatxle units were In use on the Soviet Russian-gauge railways at any given time The figure excludes cars on narrow-gauge lines belonging to the Timber and other minlsterles and not under tho Transportation Ministry and alsoelatively small number of Russian-gauge cars owned by industrial enterprises and not at the timehe general railway net, as well as cars used for some railway operations. Error through such exclusion is not likely toercent. No correction for this has been made, because the Soviet statistics under analysts exclude these items, and inclusion here would dcstry consistency between the various figures.

In reducingotal to actual cars, assumptions must be msde as to the relative proportionsxlexle cars employed. We have Beshchev's statement that "half Of the total car park isxleut the total park obviously includes cars laid up for repairs or laid up in reserve, as well as

the fleet actually in use, andxle cars are on the average much never and in greater demand thanxle ones. Another Soviet announcement2 says that "more than half the freight cars in/ One remark applies to total parknd the other to fleet in use Consideration must also be given to the fact that freight cars built lo the latter part23 will havexle units, while retirements will have beenxle ones, so that the fleetse will be muchxle one than it vas2 when Oudok commented. Direct observation of Soviet trains34 Indicates that as of recent months aboutoercent of the activexle cars, and, for reasons already given, an average ofercentxle cars seems reasonable

Assumingercent of the active fleet toxle units, totalsxle carsxle cars in operation at any given moment are'derived,ombined total in physical unitsars ln use.

It is of interest to note that the figure for cars ln use at any one time has increaseduch smaller rate than has tha figure for tons of cargo originated. Decreased turnaround timeIncreasing the tonnage carried per car per yearaccounts for most of the difference, but the Increased percentagexle cars is also significant, sincexle car actually carries morexle ones. Heavier loading per car Isactor.

Estimated Total Freight Car Park.

The speeches contain little data useful for deriving the total Trelght car park, even though much material on cars Is presented. Nevertheless, it seems possible to give rough When Beshchev says that half the park isxle units, he probably means some figure in theoercent range. igure ofercent will be here assumed. He also stated that there are delays in effecting repairs, and this may imply larger than average layup of cars.

Assuming an equal numberxlexle cars in the total park, andercent addition to the numberxle cars in the operating fleet to allow for cars laid up for repairs, andercentmall inactiveotal li-axle park may be derived as follows:


4-axle cars In

It-axle cars laid up for repairs or In re nerve 0


On the assumption based on Beshchev's speech that there is an equal numberxle units, there must alsoxle corn in exijtence. If the figurexlc cars, which has already been obtained for the active fleet. Is subtractedotal is derivedxle cars laid up for repair or for an emergency reserve, with most of the total probably In the reserve category. It can be seen that the emergency reserveIn all probability mostlyxle cars, and this conclusion la not affected by any reasonable increasexle cars assumed laid up ln such reserve status, since in order to preserve theoatio in total car numbers, one must assume an addition of equal numbersxle cars to counter similar numbers ofxle ones so allocated.

Expressed Inxle unite, the total Soviet freight car park may be of the orderars, made up of an active fleetars and ma Inactive onears. It la unlikely that the parknits, since this would imply an inactive fleetxle carsxle cars,otalxle units. This would give about one-third the total car park as laid up for repairs oreserve, and by Western standards this vould be excessive. However, it la known lhat some sort of reserve of cars ln good condition does exist.

If the ratioxlexle cars ln the total park Isoercent, the minimum number of cars may possibly be as lowxle units.

The above figuresotal freight car park that might be up toercent smaller than was givenreviousxle units. These rigurea alsoather low rate of growth In theears. This is not Inconsistent with recent Increases in tons carried and merely means that additions ofxle cars may havc been almost balanced by the scrapping ofxle cars fonatrly carried on the inventory but of doubtful value. It is known that as7


the cax Inventory vas Inflated with much la id-up stock that had suffered war damage, and much of this nay have been scrapped rather than repaired.

In connection with the computation of the Soviet freight car park and its allocation as between active and reserve fleets, attention is called to the fact that it is not known whether Beshchev was considering the total Soviet Russian-gauge figure or only the cars owned by the Soviet railroads, when he was discussing car supply. Some Russian-gauge cars are owned by the steel and other industries. Part of the total listed asn the above tabulation, therefore, really might be cars in use around steel mills or mines, but not moving on the tracks under railroad Jurisdiction. Since by its derivation the "active" figure definitely excludes cars not under railroad Jurisdiction and the total figure derived from consideration of percentages may be for toe entire USSR and not for the railroadsubtraction of "active" from "total" may not give railroad-owned inactive,ixed residual. The error cannot, however, be large.

13- Relative Importance of Rail Transport Compared with Other Methods of Transport.

For many years Soviet rail traffic has produced aboutent of the annual total ton-kilometers, as compared with aboutercent furnished by water transport. Plans have been made repeatedly to shift part of the movement of bulk commodities to river or sea transport, but the plane never seem to materialize. Kaganovich again proposes to Induce greater use of combination rail-water movements, but at the same time Beshchev and others promoting combinationfurnish data which show that for much of Soviet industry the rate advantages obtainable are minor, while the delays in getting goods may be great. It is not thought that the campaign toarger percentage of traffic to the waterways will succeed unless rate savings are made more attractive, and penalties for delayed industrial output are reduced.

Particular attention is given in the speeches to cotton movement, which was once taken by the Caspian and Volga and now is essentially all-rail, and to timber hauling on railroads paralleling the Volga.


No mention whatever ie made of pipeline transport in the speeches, aad truck transport receives only moderate attention. It is, however, stated that truck movement is scheduled for increast and that trucks now are idle an undesirable proportion of the time.

The Kaganovlch speech shows thatraffic Inwas divided between various media as follows: percent;ercent;ercent;

Container Traffic .

ess than carload freight increasedercent according to Beshchev, and he implies that the major growth is still to come. Kaganovlch lays great stress on promoting larger movements of consumer goods and in discussing the future says. It is necessary that ve should haveontainers as at present,illionillion containers." Containers are very commonly used to carry less than carload freight.

reat expansionovement of smallItems is planned, much of which will be consumer goods. Along with this go plans for Increasing the supply of refrigerator cars and for speeding their turnaround, so that the Soviet public may get more meat, fresh fish, butter, fruite, and vegetables.

Quite clearly, an increase In the Soviet standard of living is intended, or at least promised, and this may indicateeed or desirability to conciliate the public and show results from Communism,ack of plans for major aggression in the next few years. It is emphasized, however, that consumer goods must not move at the expense of basic raw materials. Kaganovlch states that preferential categories In goods movement have been abolished and that all commodities will in future be treated alike as equally entitled to transport. Then, after devoting much time to showing how movementcof consumer goods vill in the future be fostered, he warns the railroad men that if they io not move the coal, iron, and other key items at the expense of lesser Items, there will be trouble. In other words, the categories of preferred and secondary items are aboLished In name andin fact.

15- Norms.

A study of Soviet press items indicates that norms for weight of individual trains operated vary from region to region, and that they are also changed from time to timeof course, usually upward.

For example, it is stated that the Baltic Railroad3 carloading plan was completed byecember;0 above-norm-weight trains were operated, haulingillion tons of freight above the norm, and that to haul this above-norm freight anrains would (otherwise) have been needed, y

This works outorm for freight ofons per train and Indicates that the above-norm trains of this railroadons of freight Norma for the Baltic area could be expected to be lower than the"Soviet average, since mining and heavy industry are not prominent.

Another article, datedgain3 rail operations for the Baltic Railroad, stated that to haul the freight (of unspecified amount) carried above norm, anrains of normal weight would have had to be .dispatched. 8/ Allowing for probable carriage fromhroughecember, this wouldorm for freight per train orons. Between these two dates, the norm for the Baltic Railroad was probably raised, since Beshchev stated that norms for Soviet trains had been raisedU.

It might be reasonable to assume that someone writing the news stories asked "how many normal trains would the above-norm weight freight havend that in December the figure was worked out using the then-current norm, while in January the new one was used, without realization that it would have beento past performance. An alternative theoryecomputation3 performance, adding tnays,uch smaller total as carried by above-norm trains, seems less Likely. Such an interpretation is partly borne out by the fact that the saving through operation of above-norm trains is quoted asillion rubles" in both articles, which might imply that the tonnage carried above norm was about the same in both computations. Topercent reduced train number with the same norm for carriage, the tons


carried would have had lo De reducedercent on adding in carriage for the additional days and recomputing. But this would have changed the saving made, soifferent figure for this could have been expected.

The third alternativethat the figures are mere fictionseems unlikely in view of cross checks that have been made oo other Soviet railroad data. These checks seem to show that information published is usually accurate, although it may not be complete.

IV. General Comments3 Operations.


The. speeches under analysis give much data on locomotives, but it has not been possible as yet to get present locomotive numbers or average doily runs therefrom. It seems Clear, however, that in spiteampaigns to get more work out of engines and to haul larger trains with fewer locomotive numbers, the engine supply is basically adequate and the average age of equipment low. Kaganovich states thatercent of the work is being done by locomotives built Blazhenov states -that thc daily productivity of US locomotives Is onlyercent of .those of the USSR. 9/ Beshchev soys that "at the present time, the locomotive park is significantly larger than pre-war., Comments on the plan for building more powerful locomotives to handle heavy trains have already been given.


Fuel consumption and operating cost data are found, but inittle too ambiguous to admit of useful calculation. Some items follow:

the weight of the train by only 5the capacityectionercent and decreasesthe same amount the demand upon the locomotivereduces fuel expenditure by not less '

the weight of freight trains be raised bypercent all over the country, the annual exploitation expenses

of the railroads would be reduced by aboutillion rubles. Should the run of every locomotive increase byilometers per day the annual economy would amountillion rubles. Heavier '


trains mean also an economy of fuel. An increase in the weight of train loads byercent vculdorresponding increase of fuel consumption ofl/

c. "Although the average daily freight locomotive run3 increased as compared3 plan was not Fuel consumption per ton-kilometer wasercent as compared

The average length of haul by commodities3 was reported to have been as follovs: ilometers;ilometers; ferrousilometers;ilometers; and mineral buildingilometers. All of these are Items for which length of haul is cited as having Increased to an excessive degree. Statistics are not given for coal and petroleum, so it may be.inferred that they either met or only slightly exceeded planned lengths of haul. Nevertheless, for coal, complaint is made by Kaganovlch that cross hauls or those of undesirable length amounted toillion tons andillion rubles. Thirty percent of all coal was hauledilometersorcentilometors. Kuznetsk coal went as far an Moscow.

The coat of shipping petroleum products over the Trans-Siberian Railroad to the Far East was moretlUon rubles

From rate data at hand, this would imply shipment ofillion tons of petroleum products. Some data on movements of crude oil and fuel oil in the Volga region are given by Kaganovlch, and it may be possible In the future to work out freight flows


6 percent of freight cars had automatic brakes. Beshchev states that nowercent are so equipped. Automatic couplings were5 percent of the freight cars Beshche-shows that the figure is5 percent. Failure of car couplings while trains are moving are apparently still frequent, however, to judgeemark of Kaganovlch.

It has been stated thatesult of the operations of above-norm-weight trains, the average weightreight train has grown3 byercent In comparison The same source says that Innlyercent of all freight trains were above-norm weight; the others were of normal weight Or

Material oo railroad finances and on labor Ls given by Kaganovich and Beshchev, butittle too general for useful statistical analysis. It is, however, clear that the railroads were profitable3 and that the efficiency of labor lsigher proportion of the labor force seems to be employed io rail transport io the DSSR than would correspond to US practice. Much is made of housing developments for railroad employees, but inspection of the claims shows that, even if true, they represent almost nothing per worker.

It is planned that capital investmentU will6 billion rubles for theillion rubles for river and sea navigation,illion rubles for truck transport. No mention was made of air transport or pipeline plans. These figures are higher than3 andelative increase in al locations to waterborne' movement, ln line with Kaganovich's renewed demand for shifting some of the burden of bulk freight movement from rail to water-rail hauls.

Electrification is scheduled0 kilometers of lineentering mainly in the Volga-Urals-Weet Siberia area.

The general tone of the speeches, the information furnished, and the inferences which it has been possible to draw, combine to give the impression3 Soviet railroad performance wasbe whole satisfactory, although operations wereevel whichstrained the system. "Bolehovik self-criticism" was much in evidence in the speeches, but seemed to be the normal Communist nagging and slave-driving by higher officials, without any overtones of real dissatisfaction with past performance. It is clear, however, that the April-May meetings were considered more Important than those in other recent years. The Inference seems roasonable that the meetings were not so much to Jackailing Industry as to serve notice that drastic changes lo freight volume and character would soon take place, and that these would require heavier and faster trains running oo tighter schedules, and with reduced Urns in yards and terminals. The railroads were warned about this, and means for meeting the problem were pointed out. An undertone of basicthat the problem could be solved Is discernible. It is believed that although the planned railroad traffic expansion will seriously tax Soviet ability over the next few years, the problem will be successfully met.


V. Approximate Tonnage and Value of Gocas Transported, by

In thc Kagauovich speech ofortaln statisticsabout the total value of Soviet production, plus someon tonnages of the important bulk commodities handled bv

Kaganovlch says "the lowering of Industrial production coat byercentaving of moreillionhis implies an industrial production cost ofillion rubles.

He also states that "expenditures Incurred by ministries on goods transport3 amounted to aboutillion rubles, of1 billion waa for rail transport."

There la also the additional statement that "at any given moment


igure Of

illion tons of goods shipped per year has already been derived from the statement,alueubles per ton is eaailywlet raUts3 must then be

UUnreported cost of

Industrial production. Some explanation seems necessary.


thCrethrough carriage and

counting of the same manufactured items or their canponent parts several times. This duplicationajor factor. Also, part of the difference, of course, can be accounted for by the value of raw materials and foodstuffs hauled. As will be later seen, however "his

nay be the Soviet write-up of manufactured goods coats toenerous profit margin before sale. That Is, the comparison made here onmOQufnctures alue of items moved

of cl-^fylng the problem Is to list the principal bulk

commodities that were not likely to have been reshipped to any We degree; to give what can be found about total tons probably moved unit value, and total value; and to subtract total tons and total value from figures for all commodities combined that are listed above, and then to study the residuals.


Too following table" gives thc present best estimates of the tonnages and values of the major bulk commodities moved It will be nnted that, by this coapuUtion, the bulk ccantodlties are nearlyercent of the total tonnage movedhe Soviet railroads. Industrial bulk commodities arepercent, aad bulkitemsercent of the movement. For Industrial rawheck is possible, since Krlvomelko states that bulk Industrial commodities make upercent of the total railway For agricultural items, an article states that tho proportion that agricultural products bear to the total rail traffic has decreased3 hut is now atill overercent, and that the actual volume of agricultural Items shipped has risenercent Agricultural items as defined in this article include flour, meats, vegetables, and other Items that are tabulated in tho Table as part of the "manufactures" class. The calculation, therefore, that the enumerated agricultural bulk ccmraoditles sake up only 6of the traffic la quite conaiateot with the fact that allitems combined wereercent of3 total.

VI- Total Freight Revenue3 and Average Freight Revenue

Kaganovich states io hlapril epeech that "expenditureministries on goods transport3 amounted tof3 billloo was for rail transport.- Railroadfrom other than government traffic should be small sinceprobably consist mainly of minor movements of privateshipment by cooperatives, and that part of the overlandChina and the European Satellites that may be carriedSoviet Intervention as

ough approximation, railroad freight revenue might be about h? billion rubles. Then,.since freight carriage amountedillion ton-kilometers, the average revenue per ton-kilometer might have amountedinimum ofopecks.


In view of thc incomplete nature of the data oow onefinite finding cannot at this time be made as to the significance ofopeck figure. . Vinnlcbenko gives the percentages of rail-road revenue planned to be contributed0 by all the important bulk commodities. From thla It la fc-ind that "other" items (mainly

The table follows on


E.tlcmted Volume mod Value of Bulk Conxodltie* Originated on Soritt Railroad*


(Metric (Rubles)

Props, Crosstlea, etc.)



-aval. Stone, Flux, Brick)



0ravel, ate.


Dtnote for the Table follows on



Estimated Volume and Value of Bulk Coamodltlei Originated on Soviet Railroads a/


Amount Value perValue

(Metric (Rubles)



Ginned 0


Mineral Fertilizers


Total Bulk Commodities

Average Value per

Other Commodities (Manufactures and Processed


Aversge Value per

Total Commodities

a. Coal and timber estimates are derived from data given by Kaganovlch. Estimates of other other commodities are those of CIA commodity analysts.


manufactures) at that time were scheduled to3 percent of the revenue total. It Is also found that average freight revenue was planned to0 kopecks. But two major rate reductions took place0utting freight ratesotal or aboutercent, so that If the composition of freight traffic vas the same3 as It vas planned to behe average rate3 ought to have been2 kopecks per It la thought that there was en an averageery slight differential reduction In favor of low-rate bulk commodities, so it appears that the composition of freight traffic has changed and that manufactured Items and processed foodstuffs may now moke upercent of the railroad revenue total. That is, the railroad revenues give evidenceeven If at present only tentativelythat the value of the output of manufactured Items ln the USSR has increaseduch greater rate in theears than has that of crude materials, even though the latter nas grown very substantially. It also seems likely, from thc average value of manufactures per ton shown in the preceding table, that the manufactures now moving are of rather highly processed types. These findings tend to bear out Soviet claims regarding the growth of industry.

Verification of the basic reliability of at least uomo Soviet statistics and economic claims is one of the more important products of this memorandum.






Moat of the information ln this memorandum is based on speeches by Lnznr Kaganovich. Beshchev (see Foreword). All references in the text to these speeches are documented aa follows:

Uxor Kaganovich, "Speech before the Supremeoviet* (FBIS Dally Report

I.azar Kaganovich, "Speech before the RailroadovietFBIS Dally Report No/lO^

Beshchev,eported in Cudok,prreported inay

to those speeches in this memoran-dum Individual source citations for each reference to them wUl not

ill be given In JhU

Evaluations, following the classification entry anahave toe following

of Information

- Documentary Completelysuallyairlyot usuallyotannot be juLjid

" Confirmed by other sources

Probably true

Possibly true

Doubtful Probably false

annot be judged

* SS

not otherwise designated are those appearing on the .ted document; those designated "RR" are by the author of this Ho "RR" evaluation is given when the author agrees vith the aluation on the cited document.

Gudok, U. Eval. RR 2.

H. Hunter, Soviet Transportation Policy, Haverford College.

U. Eval. RR 2.

' 3-

' k.

U. Eval. RR 3-

U. Eval. RR 2.

Sovetakaya Latvlya, U. Eval. ItH 2.


BluKhenov,BIS Dally Report3

U. Eval. RR 2.

U. iKval. RR 3.

Blazhenov, op. cit.

U. Eval. RR6 U. Eval. RR 2.

Krlvornelko, Organization of Railroad Car

U. Eval. RR 2.

Ekonoalkl, Ko. U. Eval. RR ?.

yi -

Original document.

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