R.W.Davies, M.B.Tauger and S.G.Wheatcroft
Most western and all Soviet studies of the stalinist economy have ignored the role played by the stockpiling of grain in the agricultural crisis of the early 1930s. Thus in his major work on stalinist agriculture published in 1949, Naum Jasny frankly admitted that data were insufficient to reach a conclusion, merely noting that "stocks from former years probably declined during 1932"2. Baykov, Dobb, Volin and Nove said nothing about grain stocks.3 At the time, western commentators
did pay some attention to the possibility that the stockpiling of grain exacerbated the famine. In autumn 1931 Japan invaded Manchuria, and in spring 1932 British diplomats reported that Karl Radek had told them that, owing to the threat of war in the far east, enough grain had been stored to supply the army for one year.4 In February 1933 the notorious but shrewd journalist Walter Duranty wrote in The New York Times of "the unexpected additional demand for grain necessitated by the Far Eastern war danger last winter".5 Since the food and fodder grain consumed by the Red Army in one year amounted to about 800,000 tons,6 this would have been enough to provide a rather modest annual bread ration for several million people. A stockpile of this size was, of course, less important than the 4.79 million tons exported from the 1931 harvest or
even than the 1.61 million tons exported from the 1932 harvest (see table 4). But was such a military stock accumulated in those years? Enlightenment had to await the opening of the Russian archives. The impact of the first revelations about grain stocks has been dramatic. On the basis of a preliminary, unpublished typescript by the eminent Russian historian V.P.Danilov, Robert Conquest has announced that the archives have revealed that in the famine year of 1932-1933 Stalin was holding immense grain stocks, the existence of which was previously completely unknown. He wrote in this journal:
there were 4.53 million tons of grain in various reserves - the 'Neprikosvennyi Fond' and the special 'Gosudarstvennyi Fond', neither (he [Danilov] points out) justified by any danger to the country, and readily available to prevent the real danger - mass death by famine.7
Addressing a wider public in The Times Literary Supplement Conquest further explained: even apart from the fact that the 1.8 million tons of grain exported would have been enough to have prevented the famine, there were in addition two secret
grain reserves between them holding 4.53 million tons more, which were not released to the starving peasantry.
Grain stocks of 4.53 million tons would certainly have been enough to feed millions of peasants in 1932-1933. One ton of grain provided a good bread ration for three persons for a year, so 4.53 tons would haveprovided bread for some 13-14 million persons for a year. In view of the importance of grain stocks to understanding the famine, we have searched Russian archives for evidence of Soviet planned and actual grain stocks in the early 1930s. Our main sources were the Politburo protocols,
including the osobye papki ("special files" - the highest secrecy level), and the papers of the agricultural collections committee Komzag, of the committee on commodity funds, and of Sovnarkom. The Sovnarkom records include telegrams and correspondence of Kuibyshev, who was head of Gosplan, head of Komzag and the committee on reserves, and one of the
deputy chairs of Sovnarkom at that time. We have not obtained access to the Politburo working papers in the Presidential Archive, to the files of the committee on reserves, or to the relevant files in military archives. But we have found enough information to be confident that this very high figure for grain stocks is wrong and that Stalin did not have
under his control huge amounts of grain which could easily have been used to eliminate the famine. The definition of "grain stocks" is a complicated business. The literature divides them into two main categories: "invisible stocks" (nevidimye zapasy) and "visible stocks" (vidimye zapasy). The former are those held by peasants (and in the 1930s by collective and state farms) for food, seed, fodder and emergencies. Peasant carry-over is very difficult to assess; the official estimate for 1 July 1926 was 7.21 million tons, while a careful independent estimate amounted to only 4.19 million tons.9 These
calculations were of some politico-economic importance: the central political authorities believed and sought to demonstrate that peasants and collective farms were concealing substantial stocks; peasants and collective farms sought to minimize knowledge of their stocks. During the grim winter of 1932-1933, the authorities seized the seed stocks of collective farms on the pretext or belief that concealed grain stocks were available to them. In the archives widely-varying estimates of
invisible stocks for the early 1930s may be found; not surprisingly, they show a general decline in the course of 1931-1933, and an increase in following years.10 The "visible stocks" rather than the invisible stocks will be our main concern in this article. These were those which had passed from producers to traders, to state and other collection agencies and to subsequent grain-consuming organizations, plus the stocks in transit. Soviet statistical agencies estimated the total of
all visible stocks on l July 1929 at 1.76 million tons)11, of which
there were:
held by state and cooperative collection agencies .912
held by consuming organisations (including industry).331
miscellaneous .141
in transport system .376
The figure in Table 1 for 1 July 1929, 781,000 tons, is a revised official estimate by Komzag of the figure given above for state and cooperative collection agencies, 912,000 tons. It thus excludes grain held by consuming organisations and in the transport system. This was that part of the visible stocks which the state had more or less readily at its disposal for distribution to the population, for export and for other uses. These stocks were generally known in the statistics as
"availability with the planning organisations" (nalichie u planovykh organizatsii); we shall refer to them here as "planners' stocks". Planners' stocks were further divided into "commercially available" (kommercheskoe nalichie) and "various funds" (raznye fondy) (see table 2). The "funds" were those parts of the planners' stocks which were set aside for special purposes, sometimes in special stores, sometimes merely notionally. As we shall show, the funds included both the "untouchable fund" ("Neprikosnovennyi fond" or "Nepfond") and the "mobilization fund", also known as the "state fund" "Mobfond","gosudarstvennyi fond" or "gosfond"). "Commercially available" was something of a misnomer: it referred to stocks held by grain- collection and related agencies which could be passed on to consumers in accordance with an approved plan of utilization. Grain stocks naturally varied considerably during the course of the agricultural year, reaching a peak immediately after a harvest, and falling to their lowest levels just before the next harvest. Harvesting and the grain collections began in the south in early July, but in many other areas not until August. Normally the 1 July figure was given as the minimum level of stocks; but
this was not quite accurate. During July, grain available from the new harvest in the month as a whole is less than grain consumed, and stocks continue to fall until the last days of the month. 1 August would be a better date for assessing minimum stocks, but data for that date are not always available. Thus, quite apart from the need for a permanent grain
reserve, a major problem for the central authorities was the need for "transitional stocks" (usually known as "perekhodiashchie ostatki" to enable continuous supply at the end of one agricultural year and the beginning of the next. Ever larger transitional stocks were needed from 1928 onwards, with the attenuation of the grain market and the
dependence of larger numbers of people on state supplies (including many peasants in grain-deficit areas). From 1930 onwards state allocations of grain for internal purposes only (food rations, army, industry, etc. but excluding exports) amounted to some 1.35-1.5 million tons a month. Moreover, areas requiring supplies were often thousands of kilometres from the main grain-producing areas; and, once available, the grain had to be processed and delivered. In the course of establishing a state grain monopoly in the mid-1920s, the Soviet authorities did not succeed in building up a state grain reserve. In December 1927 the directives for the five-year plan approved by the XV Party Congress stressed the importance of the accumulation of stocks in kind and foreign currency reserves during the course of the plan. The accumulation of stocks of
all kinds would achieve "the necessary insurance against large vacillations in the conjuncture of the international market, and against a potential partial or general economic and financial blockade, against a bad harvest within the country, and against a direct armed attack."12 But a Soviet grain handbook published in 1932 noted that "all attempts to create a large grain reserve did not have positive results", even though "the difficulties experienced in 1927/28 and 1928/29 revealed the
categorical necessity of creating such a reserve".13 According to Soviet data, on 1 July 1929 the total amount held in the state grain fund (gosfond) including the remnants of the centralised milling levy from the previous harvest, amounted to only 69,000 tons.14
The 1929 harvest and the 1929/30 agricultural year.
On 27 June 1929 the Politburo adopted a much-increased plan for grain
collection from the 1929 harvest, resolving:
In accordance with the resolution of the XV Congress on the formation
of a grain fund, it is considered necessary to create an untouchable
stock amounting to 100 million poods [1.638 million tons] of food grains... It is
considered that the untouchable stock may not be expended
by anyone in any circumstances without special permission
from the Politburo and Sovnarkom of the USSR.15
Two months later, on 29 August 1929, Stalin wrote to Molotov, praising the success of the first stage of grain collection from the 1929 harvest and emphasizing the importance of reserve stocks, that "we must and can accumulate 100 mln poods [1,638 million tons] of untouchable stocks [neprikosnovennye zapasy], if we are really Bolsheviks and not empty chatterers."16 By the beginning of December, 13.5 million tons of grain had been collected, well over twice as much as on that date in any
previous year; and the first drive for the collectivization of agriculture was rapidly accelerating. Stalin, jubilant and jovial, again wrote to Molotov: "Greetings to Molotshtein!... The grain collections progress. Today we decided to increase the untouchable fund of food grains to 120 million poods [1.966 million tons]. We will raise the rations in industrial towns such as Ivanovo-Voznesensk, Kharkov, etc."17 The grain handbook of 1932 noted that the establishment of a grain
reserve "was posed as a central and top-priority task for the grain campaign of 1929/30".18 The main statistical journal, reporting record grain stocks accumulated by January 1, 1930, noted that "a fundamental difference between the stocks of the current year and the stocks of the previous year is the formation of a special fund, not used for current needs, while in past years grain was used entirely for meeting current requirements." The journal described this "untouchable fund" as "having an insurance function in case of a bad harvest or any other extraordinary needs."19 Although grain collections from the 1929 harvest were extremely large, they had both to supply grain to many consumers who had previously obtained it on the peasant market and to provide for increased export. Planners' stocks increased by 1.3 million tons between 1 July 1929 and 1 July 1930, reaching 2.084 million tons. The Politburo deemed it possible to allocate only 786,000 tons of this to the Nepfond
on 1 July 1930; but explained that this amount was to be "absolutely untouchable."20 The 1930 Harvest and the 1930/31 Agricultural Year. The harvest of 1930 was exceptionally good: collections were 38 percenthigher than in the previous year, and more than twice as much as in 1928/29 (see table 4). Planners' stocks on the peak date of 1 January 1931 were even higher than on 1 January 1930 (see table 1); on that basis the Politburo concluded on 7 January that the Nepfond could amount
to 150 million poods (2.457 million tons) and that, in addition, the "mobilisation fund" (Mobfond) could amount to 50 million poods (.819 million tons) - 3.276 million tons in all.21 The Mobfond was later described by Kuibyshev as intended to provide adequate grain (and other commodities), amounting to 1_-2 months' supply, to cover delays in supplies during mobilization, and also to make some provision for the largest industrial and political centers.22 But, although planners' stocks had increased to 2.332 million tons on 1 July 1931 and remained as high as 2.026 million tons on 1 August 1931 (see table 1), they were far below the level of reserve stocks proposed by the Politburo on 7 January 1931.
The 1931 Harvest and the 1931/32 Agricultural Year.
Unlike the 1930 harvest, the 1931 harvest was poor (and much worse than the Soviet political authorities believed). Nevertheless grain collections in the agricultural year 1931/32 slightly exceeded the 1930/31 level (see table 4), and the authorities continued their efforts to accumulate substantial reserve stocks. Their aims were now somewhat less ambitious: in October 1931 the Politburo decided that the Nepfond and Mobfond together should total 150 million poods (2.457 million tons), as compared with the 200 million poods specified in the Politburo decision of 7 January 1931. But it also decided to consolidate central control over the reserves: both the "grain Nepfond and the grain-fodder Mobfond" were to be transferred from Narkomsnab (the People's Commissariat for Supplies) to the committee on reserves 23 - a powerful organization, whose chair was Kuibyshev and whose vice-chair, Iagoda, was head of the OGPU.24 The Politburo intended that "warehouses and
personnel" should also be transferred to the committee on reserves; but at this time they apparently remained in the grain collection and processing network. Use of grain deemed to be part of the Nepfond or Mobfond required permission of the committee on reserves or even the Politburo. Sovnarkom further decreed that all 2.457 million tons were to be transferred to the committee on reserves by 1 December 1931, togetherwith large stocks of other foodstuffs, consumer goods and metals.25 By 1 January 1932, the grain set aside in "various funds", nearly all ofwhich was Nepfond and Mobfond, amounted to 2.033 million tons (see table 1): the plan for the reserve funds had been largely achieved. But this apparent triumph was short-lived. The demand for grain relentlessly increased. Grain exports in the agricultural year 1931/32 were one million tons less than in 1930/31; simultaneously, however, state grain allocations within the USSR increased (see table 4). The increase in
internal utilization in 1931/32 was part of a process which had been proceeding relentlessly since 1929, resulting from: a substantial increase in the number of industrial and building workers and their dependents; a growing necessity to supply grain for seed and food to collective farmers and others in areas where harvests had been low and grain collections too high; an increase in the use of grain to feed sections of the population, including cotton-growers and timber-cutters, who had previously obtained their grain from the market, and to feed exiled kulaks and others; an increased consumption of grain by industry. The total amount of grain allocated by the state for internal useincreased from 9.015 million tons in 1928/29 to 16.309 million tons in 1931/32; in 1931/32 alone the increase amounted to 2.477 million tons. 26 The pressure on stocks was relentless. Despite demand, the Politburo endeavored to reduce the rate of issue of grain. In March 1932 it agreed to drastic cuts in the bread ration for consumers on the lower-priority ration Lists 2 and 3 27. Many requests for additional rations, even from high-priority industries, were refused. These reductions, and the irregular delivery of bread and other food supplies, led to famine in the towns in spring 1932. Among the urban population of the lower Volga region the death rate more than doubled between January and July 1932; among the urban population of the Kiev region it increased by 70 percent; and even in Moscow the death rate rose by one-third.28 But the severe measures of March 1932 failed to reduce food allocations to which the state was committed to the level of the available grain. On 23 May 1932, an alarmed Kuibyshev prepared a emorandum concerning the grain situation for the Politburo, in which he outlined the additional measures needed if an uninterrupted grain supply to the main industrial centers was to be maintained until the new harvest; his proposals even included the reduction of the bread ration for workers on the Special List and List 1. The draft memorandum preserved in the Kuibyshev papers includes his handwritten note in blue crayon:
With a full sense of responsibility I want to
emphasize that last year we had 88.8 million poods [1.45
million tons] [of food grains] on 1 July, and that in the
current year there will be only 57.7 million poods [0.945
million tons].
What does this mean?
It means that we can cope with the supply of
bread only by an exceptional level of extremely
thorough organization.
Another handwritten sentence, crossed out, reads, "I ask you to give tothe committee on reserves dictatorial powers until the new harvest."29The Politburo did not accept Kuibyshev's proposal to reduce rations for the Special List and List 1; but on 25 May it decided that it was necessary before 1 July to collect the outstanding 14 million poods [229,000 tons] of grain from the remains of the 1931 harvest, to add more barley to the food grains and to transfer various grain stocks from one part of the country to another. It also reduced the allocation to the military by about 16 percent, and called for the acceleration of the import of grain from Persia and its immediate transfer to the far east. In spite of all these measures, it was estimated that planners' stocks of food grains (excluding fodder) would decline from 2.01 million tons on 10 May to .886 million tons on 1 July. For the difficult months of July and August 1932 when the new harvest was beginning to come in, the Politburo also resolved that all grain collected from the new harvest would be used solely to supply industrial centers and the army.30 In the outcome, planners' stocks on 1 July 1932 were as low as the Politburo had anticipated in May: food grains amounted to 915 thousand tons and all grains to 1.36 million tons - 1 million tons less than on 1 July 1931, and even less than on 1 July 1930. The Nepfond and Mobfond, intended to total 2.457 million tons, and reaching about 2 million tons on 1 January 1932, amounted to only .641 million tons on 1 July (see table 3). The demand for grain had impelled the Politburo to use up most of its "untouchable" fund. On 1 July total stocks of food grain amounted to about one month's supply: in Ukraine, the lower Volga and north Caucasus less than a month's supply was available.31 Following the Japanese invasion of Manchuria, the authorities had utterly failed to build up grain stocks in east Siberia and the far east: total stocks of food and fodder grains in these two regions amounted to at most 190,000 tons on 1 July;32 the 1 million tons of military stocks that Radek reported to the British was apparently sheer bluff.
The 1932 Harvest and the 1932/33 Agricultural Year.
In May 1932, in preparing its plans for the forthcoming harvest, the Politburo somewhat reduced the grain collection plan below the previous year's level, and sought to fill the gap by permitting trade in grain at market prices once collection quotas had been met. But the sharp decrease in grain stocks below the 1931 level had greatly alarmed the authorities. In spite of the reduced collection planned in May, on 16 July 1932 the Politburo again sought to set aside substantial stocks in
the Nepfond and Mobfond from the new harvest. It resolved that in 1932/33 the Gosfond (state fund, another name for Mobfond) would amount to 55 million poods (.901 million tons) and the Nepfond to 120 million poods (1.966 million tons), 2.867 million tons in total.33 On 9 December 1932, the Politburo approved a reduced plan for grain utilization in 1932/33 by which Gosfond and Nepfond would together total 1.966 million tons on 1 July 1933; together with transitional stocks, all planners'
stocks would amount to 3.699 million tons on 1 July, as compared with 1.36 million tons on the same date of the previous year. Thus the authorities certainly planned to hold very substantial stocks at the end of the 1932/33 agricultural year (if not the 4.53 million tons claimed by Robert Conquest). And on 1 January 1933, with total stocks at their seasonal peak, as much as 3.034 million tons were attributed to "various funds" (the main components of which were the Gosfond and the Nepfond)
(see table 1). The grain utilization plan for 1932/33 was built on illusion. While grain exports were again reduced, this time by 3 million tons below the previous year's level, grain collections declined by over 4 million tons (see table 4). The net decline in grain available for internal use amounted to more than 1 million tons (see table 4), collections minus export in 1932/33 versus 1931/32), and this placed an immense strain on resources, quite incompatible with the decision to
allocate 2.339 million additional tons to planners' stocks on 1 July 1933 as compared with 1 July 1932. Moreover, the grain balance of 9 December 1932 had assumed that no grain should be allocated to the countryside for seed and food, apart from earmarked allocations to cotton growing and other specialized areas. In the course of the first six months of 1933, the Politburo reluctantly, little by little, released between 1.99 million and 2.2 million tons in seed, food and fodder, primarily as allocations or "loans" to areas which had been stripped of grain by the state collectors earlier in the year.34 While neither large enough nor timely enough to prevent the devastating famine, these allocations did use up most of the Nepfond and Gosfond which had been set aside at the beginning of the year. In spring 1933, as in the previous year, leading grain officials addressed a series of urgent memoranda to the Politburo warning of shortages. In March a memorandum from Chernov to Stalin, Kaganovich, Molotov and Kuibyshev pointed out that receipts of food grain might be .5-.6 million tons less
than in the grain utilization plan of 9 December 1932, while expenditure might be .5 million tons more; the shortfall in oats and barley might amount to a further half million tons.35 A memorandum from Kleiner to Kuibyshev, prepared in February or March, concluded that on 1 July 1933 the Nepfond would be .256 million tons less than planned on December 9.36 Two or three months later, on 17 May, a telegram from Kleiner to Stalin and Kuibyshev makes it clear that the situation had drastically
deteriorated: "surpluses in the Nepfond are almost exhausted." To provide seed, food and fodder the Politburo had agreed to release 69 million poods (1.13 million tons) from the committee on reserves, so that only 100 million poods (1.638 million tons) remained in all its reserves; Kleiner asked for conditional permission to use a further 15 million poods (.246 million tons) from funds of the committee on reserves.37 Within a few weeks the situation had apparently deteriorated still further. On 4 June 1933, Chernov sent a memorandum to Stalin, Kaganovich, Molotov and Kuibyshev, setting out the results of the 1 May
inventory of the remaining grain (ostatki) in the USSR. Chernov estimated that, as a result of commitments made in May and June, all planners' stocks would total 84.7 million poods (1.392 million tons) on 1 July 1933, including food grains amounting to 63.8 million poods (1.045 million tons), a slightly larger amount than on 1 July 1932. In several places the memorandum referred to this estimate as the "transitional remainder including funds" (perekhodiashchii ostatok
vkliuchaia fondy).38 In practice, the level of grain stocks was apparently somewhat greater than Chernov and the other officials anticipated. When Chernov submitted the grain plans for the following year, 1933/34, to Stalin, Kaganovich and Molotov, on 4 July 1933, he stated, as he had a month previously, that the total transitional stock, including fondy, on 1 July 1933 was 1.392 million tons (including 1.045 million tons of food grains).39 But the grain utilization plan for
1933/34 approved a month later by the Politburo recorded the "availability" of all grains on 1 July, including fondy, as 1.825 million tons (including 1.386 million tons of food grains).40 The final official figure published in the grain yearbook was 1.997 million tons (including 1.397 million tons of food grains) (see table 1).(see table 2) We have not yet found any satisfactory explanation of the discrepancy between these three sets of figures. The planners' stocks available on 1
July 1933 certainly included enough grain to save the lives of many peasants. But they amounted not to 4.53 million tons but to less than 2 million tons, smaller than the stocks available on the same date three years previously. The alternative figures for 1 July 1933, including the published figure (1.997 million tons) certainly include both the Gosfond
and the Nepfond. Robert Conquest's confusion about the level of stocks may be due to a somewhat ambiguous passage in Chernov's memorandum dated 4 July 1933, submitting the draft grain plans for 1933/34 to the Politburo. He proposed that in 1933/34 the Nepfond should be "120 million poods [1.966 million tons], the same level as last year," while Gosfond should be "significantly increased to 72 million poods [1.179 million tons] instead of the 55 million poods [.901 million tons] of
last year".41 According to this draft, then, both fondy together would amount to 3.145 million tons. But Chernov's tables and the figures approved by the Politburo make it clear that "the same level as last year" did not mean the actual reserve stock in July 1933 but instead the stock planned in July 1932.42 In 1933/34 Nepfond and Gosfond had to be built up from existing planners' stocks. Thus the plan approved by the Politburo on 7 August 1933 fixed total grain stocks on 1 July 1934 at
3.941 million tons, including a total Gosfond and Nepfond of 2.776 million tons; the Politburo compared this with the total stocks on 1 July 1933 of only 1.825 million tons.43 The failure to establish reserve stocks at planned levels also meant that the efforts to build up grain stocks in the far east had again been unsuccessful. According to the published data, total planners' stocks in east Siberia and the far east amounted to only .147 million tons on 1 July 1932, increasing to .269
million tons on 1 July 1933;44 some additional stocks, not included in these figures, were held by the army itself. But the serious effort to build up grain stocks in the far east began not after the 1931 harvest, as Radek and Duranty claimed at the time, or even after the 1932 harvest, but only during and after the 1933 harvest. It was not until July 1933 that Chernov received an urgent commission from Stalin to create a "special defence fund" of 70 million poods (1.147 million tons)
in east Siberia and the far east. This grain stock would require extensive new grain stores, since those of the Mobfond in the far east and east Siberia had a capacity of only .143 million tons.45 How reliable were these data on grain stocks? After the civil war, during which local authorities underestimated the level of stocks, the Soviet authorities were anxious to obtain accurate and timely figures. In the early 1920s a comprehensive system was established, by which monthly estimates of grain stocks by local statistical departments were supplemented by quarterly on-site inventories. Statistical departments telegraphed regular "short summaries" to the center two weeks after each survey, followed by more detailed (and more ccurate) reports sent through the mail.46 The same system was used in 1930-1934.47 In 1928, A. Mikhailovskii, at that time the principal authority on grain statistics, claimed that the figures for the USSR which were assembled centrally from these data were "quite reliable".48 The data on grain stocks for 1932-1933, were also, in our opinion, "quite reliable". This is not to say that they should be accepted uncritically. The discrepancy between the lowest and highest figures for all planners' stocks on 1 July 1933 - 1.397 million and 1.997 million tons - dramatically illustrates this point. If the later and higher figure is correct, the additional .6 million tons of grain could have saved many lives. But this figure does not appear in any of the records we have used until some weeks after the end of the agricultural year, and it was evidently not known to the Politburo before July 1933. There were no private inventories of grain stocks kept for Stalin and his immediate entourage, separate from those regularly assembled by the normal state agencies; the figures in the top-secret files of Sovnarkom, of Kuibyshev's secretariat and in the special files(osobye papki) of the Politburo all coincide. These figures
also agree - somewhat to our surprise - with the figures for grain stocks published in the unclassified small-circulation Ezhegodnik khlebooborota.49 But the relationship between public and secret statistics in the USSR was complicated. While they were identical for grain stocks, the exaggeration in the published figures for the grain harvests is well known. And our research shows that those responsible for planning and recording grain allocations did not contradict - even
in private - the distorted official harvest figures; they therefore included in the grain balances a large residual item - entitled neuviazka! - so that consumption could be brought in line with the alleged harvests.50 And in the extreme case of the defense budget for 1931-1934, the large increases in these years were concealed by the deliberate decision of the Politburo to publish falsified figures. The true figures appeared only in documents classified as top-secret, and
were more than treble the size of the published figures.51 The complicated relations between archival and published data can only be established by investigating each case on its merits.
We therefore conclude:
1. All planners' stocks - the two secret grain reserves, the Nepfond and the Mobfond or Gosfond, together with "transitional stocks" held by grain organizations - amounted on 1 July 1933 to less than 2 million tons (1.997 million tons, according to the highest official figure). Persistent efforts of Stalin and the Politburo to establish firm and inviolable grain reserves (in addition to "transitional stocks"), amounting to 2 or 3 million tons or more were almost completely unsuccessful. In both January-June 1932 and January-June 1933 the Politburo had to allow "untouchable" grain stocks set aside at the
beginning of each year to be used to meet food and fodder crises. On 1 July 1933, the total amount of grain set aside in reserve grain stocks (fondy) amounted not to 4.53 million tons as Conquest claimed, but to only 1.141 million. It is not surprising that after several years during which the Politburo had failed to establish inviolable grain stocks, Kuibyshev in early 1933 recommended a "flexible approach" to the Nepfond and the Mobfond, denied that they were separate reserves and even
claimed that the flexible use of the two fondy had enabled uninterrupted grain supply in spring and summer 1932.52
2. We do not know the amount of grain which was held by grain- consuming organizations, notably the Red Army, but we suspect that these "consumers' stocks" would not change the picture substantially.
3. These findings do not, of course, free Stalin from responsibility for the famine. It is difficult, perhaps impossible, to assess the extent to which it would have been possible for Stalin to use part of the grain stocks available in spring 1933 to feed starving peasants. The state was a monopoly supplier of grain to urban areas and the army; if the reserves of this monopoly supply system - which amounted to four-six weeks' supply - were to have been drained, mass starvation, epidemics
and unrest in the towns could have resulted. Nevertheless, it seems certain that, if Stalin had risked lower levels of these reserves in spring and summer 1933, hundreds of thousands - perhaps millions - of lives could have been saved. In the slightly longer term, if he had been open about the famine, some international help would certainly have alleviated the disaster. And if he had been more far-sighted, the agricultural crisis of 1932-1933 could have been mitigated and perhaps
even avoided altogether. But Stalin was not hoarding immense grain reserves in these years. On the contrary, he had failed to reach the levels which he had been imperatively demanding since 1929.