CIA HiSTCRii'AL REVIEW PROGRAM
TITLB: Estimating The Soviet Gold Position AUTHOR: Paul R. Storm
VOLUME: 7 ISSUB:
A collection ol articles on the historical, operational, doctrinal, and theoretical aspects ol intelligence.
All statements of fact, opinion or analysis expressed in Studies in Intelligence are those of
the authors They do not necessarily reflect official positions or views of the Central Intelligence Agency or any other US Government entity, past or present. Nothing in the contents should be construed as asserting or implying US Government endorsement of an ankle's factual statements and interpretations.
Tools found toorld trade bogey down to size.
ESTIMATING THE SOVIET GOLD POSITION Paul R. Storm
dr^n C^Cy Soviet activiUes5 to
^bout statistics on tbe producUon and coMuaption of nonferrous metato and minerato in the USSR
nd rare metato
W^ntbere has been no known instance of publication of the
n the produc-
Ku1' and Aluminum. It is not
m of the Soviet
"Serves should be treated with the utmost secrecy, and these
Ten ,rom m^ high-ran w
ur^L bsolute production^
nd gold reserve figures haveen published. In the face of this almost total buyout of omcial data, anything betteruesTatheof the Soviet holdings was long considered Impossible
A meaningful assessment of the USSR's financial position, however, requireseasonably accurate value bfplaeed on ite reserves of gold. The Western estimates which have traditionally ranged fromillioneh-^nfhrrdng circle that does little to Inspire confidence ln their validity-were not good enough. Better estimates had to be made on- the basiseasoned examination of allavailable to the intelligence community. First Questionable Construction
The approach that seemed to offer the best chance ofwas to begin with fairly reliable estimates that have been made of the Csartot gold reserves as of the end0 and
then compute the changes by addition and withdrawal over the followingears. An obvious weakness of thisIs that the results depend upon the accuracy ofomponent estimates of annual production, consumption, and amies, plus those of other, irregular acquisitions andBut although the number of error* small and Urge would undoubtedly be grea^^peared reasonable to expect that those on the high aide might roughly compensate for* those on the low.
A preliminary survey of available information revealed that satisfactory estimates could be made of gold collection* fron, the population and acquisitions from foreignthe Spanish gold transferred by the Loyalist government to the USSR "for safekeeping" during the civil war and that of the Baltic and East European countries which came under Soviet control when these became Soviet Republics andInformation on Soviet sales of gold outside the Bloc was also quite good for allew years ofS1 period. Consumption, almost negligible during the earlyas easily estimated for the periodold production was left a* the major stumbling block.
The USSR had published figures on production7 and there was enough additional Information to carryut after that the ground was not so firm. Soviet announcements of quarterly and annualIncreases for thead been reported and analyzed, however, by the American Legation at Riga. Latvia These report* were studied, and with some modification! the estimates were tentatively accepted
For the01 there was almost ablank of information, andime the problem ofannual production in these years seemedButumber of false starts and some wheel-spinning, data was obtainedensitive source thatled to the development of an accurate aeries offigures for most oferiod. With this major obstacle out of the way and various minor problems
clearedentative estimate of reserves as of the end1 could be reached.
Only It seemed this estimate could hardly be right. It was far lower than any made in the past, almost unbelievably low even to those who had never takenillion guesses of Western financialillion.
Moreover, the reconstruction showed Soviet gold sales in
cent years to be considerably larger tlum current produ^
requiring the USSR to hare been drawing heavily on reserves
to finance its annual trade deficits, and such Improvidence
seemed Incredible if the reserves were really so low. _.
A reexamination of the whole construction was thus called for.hortage of several billion dollars in the reserves
figure would have to derive from systematic error in a
number of component estimatesonsiderable time; noestimate or small group could possibly account foreficiency. Only estimates of production met this criterion.umber of reasons that cannot be recounted here, the accuracy of production estimates for the period0 was established within too narrow limits to leave room for anymall discrepancy, so attention was concentrated on those of the prewar.loseof the Riga analysis covering these years showed It to be closely reasoned and the estimates apparently accurate, there were several questions that had not been adequatelywhen Its figures were tentatively accepted for this study.
The first unresolved Incongruity lay in announcements made at the time by the Chief of the Main Administration of the Gold Industry, one Serebrovskly. Serebrovskly hadthat gold production increased fromillion ounces infigure also mentioned by Stalin inillion ounces6 andillion ounceshese latter figures were approximately twice the Riga estimates for those years, and the difference cumulatedears would yield anIn reserves of about USS1 billion. Serebovskiy's claims had been disregarded on the assumption that he was either Indulging In propaganda for Western ears or exaggerating for his own ends, as Soviet managers have been known to do; but now rt seemed possible that they were true.
The Dal'stroy Problem
The possible vindication of the Sercbovskiy figures would Lie ln the production ofhe only gold-producing or ganlzation not under the Main Administration of theal'stroy, the Construction Trust of the Far North, was organized by the NBTVD to make use of the horde of largely political prisoners in the middle thirties tor forced labor on the mineral resources of northeasterneria. Reports leaking out of Russia toldastregion along the Kolyma river that was rich beyond the wildest imagination. Prisoners who managed to survive the rigors of the northern winters and the tender mercies of the NKVD told of the death of millions of their fellows in theproduction of fantastic quantities of gold for the Kremlin's vaults ln Moscow.
For all their fiction-like quality, some of these reports sounded credible. One popularized tale of Dal'stroy wasolish array officer of the testimony of overrisoners, mcluding their estimates as tothe size of the labor force and the quantity of gold recovered per man. Thispututput at almostillion ounces in the year of highest production. Another account, writtenormer prisoner assignedal'stroy factory which made boxes for shipping the gold, used the quantities of boxesto calculate that moreillion ounces of gold was shipped In the peak year. Other eye-witness accountsimilar nature gave estimates of the same order. These stories had been discountedumber of reasons, but now the suspicion arose that they might be somewhere near the truth. Although, production ln Dal'stroy could hardly have matched the exaggerated guessesillion ouncesit might have reached the more conservativeillion ounces. If so, the Riga estimates obviously were low.
Doubts about the Riga reports were increased by the fact that, in spite of the sensational aspects of the Dal'stroyand the certainty that it was producing gold, they made no mention of it. Even more significant, Riga'sof production by producing area left no room for Dal'stroy, as though the analysts were not aware of the opera-
Hon or else deliberately ignored it. Most of the data used for the Riga estimates were those published by Glavzoloto, and it could be argued plausibly that Glavzoloto's production figures would not Include Dal'stroy production because Dal'stroy was not under its administration. If this was the case. Dal'stroy's production was not represented in the Riga estimates, and If DaJ'stroy's production had been very large, as large say as that of Glavzoloto, the total annual gold, production in the USSR would have been on the order of'illion ounces that Serebrovskly claimed.
' These considerationsearch for some way to establish the magnitude of Dal'stroy's, output Ins and, concurrently, for any proof as to whether the Rigawere really estimates of total Soviet productionthat of Dal'stroy or estimates of Glavzoloto's production only.
It was known that Dal'stroy's output Ins, prior to Its dismembermentad been5 million ounces annually. Finding some link between this level and the magnitude of Its output Ins wasossible approach to the determination of the latter. An intensive search was begunoviet statement comparing Dal'stroy production In the two periods.omparison, it was felt, might have been made quite innocently; there would be no reason to suspect in the USSRevelation it would be.
The search succeeded In uncovering two partial links. The firsttatement that8 the Western Directorate of the former Dal'stroy, now of Magadan Oblast. produced "not less" than it had produced in any of the previousears of its existence. The Westernas on the orderunces, roughly one-third of total output in the former Dal'stroy region in that year. Now if the Western Directorate, In accordance with thisproduced not more thanunces annually inotal Dal'stroy production ins on the orderillion ounces annually would requirein each of the four other gold-producing directorates in
DaTstroy lo have been very much greater than that in the Western Directorate, averaging morenaiion ounces each. While not impossible, this asymmetry seemed highly Improbable Every scrap of evidence available suggested that all five had occupied positions of almost equal importance In the Dal'stroy structure prior. If. on the other hand, production In the other four directorates lns bad averaged about the same as that in the Western, total pro-ductton in Dal'stroy in the peak prewar year could not" been moreillion ounces.
The second link between the thirties and fifties was found In the gross Industrial Index of Magadan Oblast, where three-quarters of the Dal'stroy gold was mined in the postwarThis index showed that the Oblast's0 was slightly greater than0 in spite of the fact that the output of large-scale industry bad remained at the same level and the outputumber of industries.timber and brick, had declinedt Is unlikely that0 gross Industrial index could have shown0 if the output of gold In Magadan had fallen significantly over the decade, particularly when that of other fairly Important Industries had declined. Production of gold constituted much toohare of Magadan's totaloutput not to aflect it.
It therefore seemed unlikely that DaTstroy's production Ins could haveillion ounces annually. The foregoing evidence, felt to be considerably stronger than tbe hearsay of prisoners who had atery limited view of the operation. Indicated that DaTstroy's major extraction areas, including the famous Kolyma, producedillion ounces In the prewar year of highest output.n ounce Dal'stroys contribution to Soviet reserves over theear period lns was thus more nearly on the order ofillionillion.
Although this conclusion leaves Serebrovsldy's claimsIt reinforces the earlier supposition that they had some other motivation than diligence in honest reporting. In retrospect, Serebrovskly's behavior opens his reliability to serious question.5 he declared that the USSR would schieve first place ln world gold outputix
months later.ovember, he said that first place could be reachedhen Justays later, onovember, he claimed that It would be reachedhe coming year. Thus In less than seven months he moved attainment of tbe goalillion ounces annually ahead four years. Either aof incredible magnitude had been discovered or behoroughlyri or frightened man That It was the bitter may be indicated bytUeear thereafter, when Serebrovskiy, along with many other senior officials of Glavxoloto. was removed from office and never heard of again. Soviet statements at the time supplemented the usual accusations of anti-state activities against these officials with specific charges of exaggeration, mentioning ln particular the practice of counting gold believed to beined but unsraelted ore.
Although Dal'slroy's peak production now appeared tono moreillionear, thethis output was included In the Riga totaloounces for the peak prewar years was still of someAgainst tbe negative evidence in Riga's failureDal'stroy and listing an "all other" category inof production not large enough tooutput. It was discovered that this distributivewashat is total production was estimatedof any area figures and then distributed,quite arbitrarily, among the various sectors. Tbetbe "all other" category was thereforealid testDaTstroy's output had been included. Moreover,Soviet announcementsincreases on
which Riga based its estimates referred, as must be supposed, to total production, Dal'slroy's output would have beenln the Riga estimates whether or not Riga was aware of It
There Is also positive evidence that Riga's estimatesDal'stroy production. An American engineer. Arthur Llttlepage, who had been Deputy Chief Engineer In Charge of Production ln Glavzoloto througheturned then to the United States and collaboratedrofessional writer in preparing an account of his years ln the USSR Not long after the book was published he died, but bis collaborator was
Interviewed In the hope that Littlepage might have left notes with him or at very least told him something about levels of production. He was unable to provide any additionalhe said that Littlepage had purposely avoidedproduction figures out of concern for the safety of his Russian colleagues, many of whom had already been arrested or were under suspicion in the purge of the gold Industry that began just after he came back. This fear of hurting hiswould have been misplaced If his published statements regarding production would have confirmed theirs, but if his testimony would have contradicted the high production claims of Serebrovskly, his concern Is understandable.
Littlepage did leave one concrete piece of evidenceemorandum of conversation describing his debriefing by members of the Federal Reserve Board records his saying that he had seen the final official plan figures for gold productionhat production did not reach 6ounces in that year, and that he did not believe it could have expanded very much in the following years, partly on account of the purges. Moreover, Littlepage at thiswas shown an article in an American mining journal which estimated the production of gold in the USSR and broke it down into Glavzoloto and Dal'stroy output Its figures were in line with the conclusions we have reached above about tha magnitude of Dal'stroy's production and with Riga's estimates of total production. Littlepage read the article and declared that it was essentially correct.
A monograph published8oviet authority on gold production, furthermore, used the same index on which the estimates in the journal article were based to show the increase In the USSR's gold production ins. This citation of the index8 is probably another coriflrmation of the article's estimates of production and, mdlrectly, of the Riga estimates: it is highly unlikely that an authority writing almostears later would use an index that reflected only one-half of Soviet output.
With the acceptance of the validity of the Riga estimates of production ins. the last serious question regarding the estimate of reserves was removed. Incredible or not, tne
analysis Indicated that1 gold holdings were short ofillion, nothing likeillion estimate still held by Western financial experts.
The experience gained In reaching this assessment does not point to the development of any standard technique orThe Important thing seemed tohoroughof all sources and pursuit of every howeverlead. Though only about five percent of the leads proved fruitful, those that paid off did so handsomely. Sources ran the gamut from the observationsakut panning for gold in one of several thousand streams in Siberia to reports from the highest levels In Moscow.
One lesson learned in the research was the unreliability of low-level eye-witness reports.mall percentage of those bearing on this problem were accurate, and thereo way, except In retrospect, of distinguishing these from themany Inaccurate ones. Published Soviet data, too, proved at times inaccurate and conflicting, although there was nothat figures put out by Soviet statistical offices were Intended to mislead.
Statements by government officials, however, were another matter. As we have said. Soviet officials have in no known Instance revealed publicly the true order of magnitude of either gold production or reserves. On the contrary: from the days of Serebrovskly to the Khrushchev visit herehen members of his entourage declared that Soviet goldamounted toillion and were being increased0 million annually, tbe consistent goal of official utterances has been to create the Image of wealth.
Yet in the realm of deeds Soviet behavior has been much more appropriateation with limited and dwindling gold reserves. The USSR has frequently foregone attractive trade offers when its efforts to obtain long-term credits failed, has lost desired deals by insisting on barter arrangements, and has been searching among its products for additional foreign exchange earners. And finally, during certain negotiations onInternational gold reserve to which each nation should contribute ten percent of national reserves. Sovietoffered, not theillion appropriate to these public claims,0 million, around ten percent of our foregoing estimate of their reserves.Original document.