The Role of Leadership Perceptions and of Intent in the Soviet Famine of 1931 – 1934

Click to increase image sizeClick to decrease image size

Notes

I am grateful for help, comments, discussion and criticism to K. Berkhoff, R. Binner, R.W. Davies, P. Ellman, M. Jansen, J. Keep, L. Viola and E. van Ree. The author alone is responsible for the interpretation and for the remaining errors.

V. Vodovozov, ‘Moe znakomstvo s Leninym’, Na chuzoi storone, vol. XII (Prague, 1925), pp. 176 – 177. Vodovozov knew Lenin personally in 1891 – 92, but was writing more than 30 years after the event in an émigré journal. A heavily edited version of Vodovozov's account of Lenin's attitude to the famine and the anti-famine NGO was included in the booklet published by the Marx – Engels – Lenin Institute, Lenin v Samare 1889 – 1893 (Moscow, 1933), pp. 98 – 101. The editor argued (pp. 98 – 99) that Vodovozov's account had ‘a particularly tendentious character’ and was quite misleading. According to the editor, Lenin did not oppose bourgeois-liberal elements feeding the hungry, organising public works etc., but did oppose seeing these activities as suitable for political exiles and revolutionary youth, as a contribution to the revolution and the overthrow of the autocracy. Lenin, according to the editor, saw these activities as a distraction from the revolution and advantageous for the ruling class since they lessened peasant dissatisfaction and despair. However, even this publication agrees that Lenin thought that feeding the starving was not appropriate for him and his comrades and was politically harmful. According to Belyakov, who did not know Lenin personally and was writing in a Soviet book in the Khrushchev era, it was not the famine which Lenin regarded as progressive but the consequences of the famine. ‘Vladimir Il'ich had the bravery to declare that the consequences of the famine [of 1891 – 92]—the growth of an industrial proletariat, this gravedigger of the bourgeois system—were progressive, because they facilitated Russian industry and brought us to our final goal, to socialism via capitalism… The famine, in destroying the peasant economy, simultaneously destroys faith not only in the Tsar but also in God and in time without doubt pushes the peasants on the path of revolution and makes the victory of the revolution easier’. (A. Belyakov, Yunost’ vozhdya (Moscow, 1960), pp. 78 – 79). The relevance of Lenin's position in 1891 in understanding the position of the Soviet leadership in 1932 – 33 was long ago argued by Conquest; see R. Conquest, The Harvest of Sorrow (London, 1986), p. 234.

Quoted from V. Kondrashin & D. Penner, Golod: 1932 – 1933 gody v sovetskoi derevne (na materialakh Povolzh'ya, Dona i Kubani) (Samara-Penza, 2002), p. 210.

R.W. Davies & S. Wheatcroft, The Years of Hunger: Soviet Agriculture, 1931 – 1933 (Basingstoke, 2004), passim. However (p. 441), Davies & Wheatcroft ‘do not at all absolve Stalin from responsibility for the famine. His policies towards the peasants were ruthless and brutal’.

Actually, the structural decision was not rapid industrialisation as such but rapid industrialisation by means of levying a tribute on the peasantry, i.e. the tribute/coercive/dictatorial model of industrialisation; see M. Ellman, Socialist Planning 2nd edn (Cambridge, 1989), pp. 96 – 110; P. Gregory, The Political Economy of Stalinism (Cambridge, 2004), chapter 2. At the time, the tribute model was opposed by the Bukharinists. Subsequently, some economists have argued that it was economically unnecessary; see H. Hunter & J. Szyrmer, Faulty Foundations. Soviet Economic Policies, 1928 – 1940 (Princeton, 1992), chapter 6; R.C. Allen, Farm to Factory (Princeton, 2003), pp. 165 – 171. This remains controversial. However, for many Bolsheviks it was politically necessary. It seems likely that a majority of the party would have endorsed Trotsky's criticism of Bukharinist policies. These policies, he wrote in 1929, might well yield fruits, but they would be ‘capitalist fruits which at no distant stage will lead to the political downfall of Soviet power’ (Byulleten’ Oppozitsii, 1929, 1 – 2, p. 22). Other models of rapid industrialisation would have had different consequences. For example, in the early 1980s China launched rapid industrialisation based on strategic integration in the world economy (the ‘open door’ model). This model of rapid industrialisation produced some results similar to, but other results very different from, the tribute model. It was the tribute model of rapid industrialisation, not rapid industrialisation as such, which contributed to the 1931 – 34 famine. Some other model of rapid industrialisation might not have done so.

The conventional view is that deviations from the trend in grain yields in this period were basically determined by the weather and the availability of traction power (mainly horses); see for example Hunter & Szyrmer, Faulty Foundations…, chapter 6. However, the cause(s) of the poor 1932 harvest is/are controversial. Tauger argued that the main cause was plant diseases such as wheat rust (M. Tauger, Natural Disaster and Human Actions in the Soviet Famine of 1931 – 1933 (Carl Beck Papers no.1506, Pittsburg, 2001)). This seems implausible for the reasons given by Davies & Wheatcroft, The Years of Hunger…, footnote 137, pp. 131 – 132. Davies & Wheatcroft (pp. 119, 128 and 439) argue that the weather was adverse, with low temperatures during the sowing period, high temperatures in the initial flowering stage, and great humidity during early flowering. D. Penner, ‘Stalin and the Ital'ianka of 1932 – 1933 in the Don Region’, Cahiers du monde russe, 39, 1 – 2, 1998, rejects poor weather as a cause of the bad harvest. She argues that there were four direct causes, a reduction in sown acreage, inadequate seed per hectare of planted land, lengthy spring sowing and the unusual number of weeds. She argues that these direct causes were a result of three shortages (of well motivated and experienced farmers, of traction power and of grain). These shortages in turn were the result of the policies of the party and the peasantry's responses to them. Penner also stresses the large harvest losses resulting from peasant attitudes. Penner's argument overlaps with that of Davies & Wheatcroft—both draw attention to the structural role of party policy, the shortage of traction power resulting from the decline in horse numbers, and the abundance of weeds. However, Penner rejects poor weather as a factor in 1932, at any rate in the North Caucasus. In her 2002 book (with Kondrashin) she extends this rejection to the Volga region. Penner relies heavily on two well-informed contemporary sources, the January 1933 report of a committee of the presidium of the all-Union TsIK, and the August 1932 report of the British-Canadian agricultural specialist Cairns, neither of which considered the weather as the cause of the bad harvest. Nevertheless, the statement by Penner & Kondrashin, Golod…, p. 424, that the 1932 – 33 famine ‘was not connected with weather conditions’ is too strong. Whatever caused the bad 1932 harvest, this statement ignores the effect of the 1931 drought on the 1931 harvest. Peacetime famines usually require two successive bad harvests.

A proposal that the regions affected by acute food shortages should be opened up to famine relief operations by international charities was made by the Ukrainian President Petrovsky in February 1932 – about a year before the peak of the famine. Had it been accepted, it might have saved a considerable number of lives. However, it seems to have got no further than Kosior, the Ukrainian party leader. It was not passed on to the leadership in Moscow. Probably Kosior thought that, given the political mood in the central party leadership, it had no chance of being accepted. However, in March 1932 Kosior did obtain for Ukraine a seed loan (mainly from the centre but also from better-off regions) of 110,000 tons; see Davies & Wheatcroft, The Years of Hunger…, pp. 113 – 114.

N. Pianciola, ‘Famine in the Steppe’, Cahiers du monde russe, 45, 1 – 2, 2004.

O.V. Khlevnyuk, 1937-i: Stalin, NKVD i sovetskoe obshchestvo (Moscow, 1992), p. 46. The same thought had earlier been expressed by Conquest. Discussing Stalin's possible role in the deaths of Kuibyshev and Gorky, Conquest wrote that ‘Nor does it seem very probable that more [evidence] will be forthcoming even when the Soviet archives are opened up. For it is rather unlikely that plans for this style of killing are committed to paper’ (R. Conquest, The Great Terror: A Reassessment (London, 1990), p. 389).

See Stalin's letter to Sholokhov of 6 May 1933, Voprosy istorii, 1994, 3, p. 22. The Stalin – Sholokhov correspondence is discussed by Davies & Wheatcroft, but their main emphasis is on Khrushchev's falsification of the whole story and the positive steps (grain deliveries, an inquiry) that Stalin took to respond to Sholokhov's account of the situation in his area. Stalin's idea that he had faced a peasant strike was not an absurd notion indicating paranoia. It seems that there really were numerous collective refusals by collective farmers to work for the collective farms in 1932; see Kondrashin & Penner, Golod…, chapter 3.

Kondrashin & Penner, Golod…, pp. 214 – 215.

R. Mucchielli, Psychologie de la publicité et de la propagande (Paris, 1970), pp. 79 – 80.

A.D. Forges, Leave None to Tell the Story (New York, 1999), pp. 66, 70, 80, 171, 227, 256 and 649.

Ibid., p. 66. The propagandist cited was familiar with Mucchielli's book. The propagandist's first proposal was to ‘create’ events to lend credence to propaganda. This of course was one of the purposes of the Stalinist show trials, which ‘created’ large-scale wrecking, sabotage and spying, which could then be used to explain economic difficulties and justify the harsh repressive measures of the state. They seem to have been quite effective in achieving these aims. For example, at the time of the Promparty trial, Gorky took the accusations and confessions at face value and wrote that ‘they [i.e. the accused in the Promparty trial] artificially created a famine in the USSR (Strana sovetov)’ (M. Gorky, ‘K rabochim i krest'yanam’, Pravda, 25 November 1930). He repeated this accusation in M. Gorky, ‘Gumanistam’, Pravda, 11 December 1930.

R.F. Baumeister, EVIL. Inside Human Cruelty and Violence (New York, 1997), pp. 43 – 45.

This was pointed out to me by Karel C. Berkhoff. Prior to this I too belonged to the ‘unintentional’ school.

Voprosy istorii 1995, 3, p. 5.

For example, it seems that the British government was involved in the assassination of Darlan in 1942; see D. Reynolds, In Command of History (London, 2004), p. 330.

C. Andrew & V. Mitrokhin, The Mitrokhin Archive (London, 1999); P. Sudoplatov, Razvedka i kreml’ (Moscow, 1996); I. Starinov, Superdiversant Stalina (Moscow, 2004); E.P. Sharapov, Naum Eitingon—karayushchii mech Stalina (St Petersburg, 2004); Yu.N. Paporov, Akademik nelegal'nykh nauk (St Petersburg, 2004).

This decree seems to have been first published in Istoricheskii arkhiv, 1992, 1, pp. 125 – 128. It was reprinted in V.N. Khaustov, V.P. Naumov & N.S. Plotnikova (compilers), Lubyanka. Stalin i glavnoe upravlenie gosbezopasnosti NKVD 1937 – 1938 (Moscow, 2004), pp. 607 – 611. The latter publication includes a number of related documents not previously published.

Commenting on the August 1931 decisions to improve conditions for the deportees, Davies & Wheatcroft, The Years of Hunger…, p. 45, rightly note that they ‘followed the pattern familiar from Stalin's “Dizzy with success” article of March 2, 1930. Economic agencies, local authorities and to some extent the OGPU itself were blamed for the inhumane consequences of the Politburo's own decisions’.

Actually the use of torture did not begin in 1937. It began earlier. For example, torture was used in the grain procurement campaign in the winter of 1932 – 33, as is known from Sholokhov's letter about it to Stalin of 4 April 1933 (Voprosy istorii, 1994, 3, pp. 7 – 18).

This well – known document was quoted in the 1956 Pospelov report (A. Artizov, Yu. Sigachev, I. Shevchuk & V. Khlopov, Reabilitatsiya: kak eto bylo, vol. 1, p. 347) and subsequently in Khrushchev's speech ‘O kul'te lichnosti i ego posledstviyakh’, at the closed session of the 20th Congress (Izvestiya TsK KPSS, 1989, 3, p. 145). It has been printed in full several times, e.g. in Iz istorii zemli tomskoi. God 1937 … (Tomsk, 1998), pp. 309 – 310, and 1936 – 1937 gg. Konveier NKVD (Tomsk and Moscow, 2004), pp. 343 – 344. Stalin was, of course, quite right about the use of torture by bourgeois intelligence services, as recent events in the ‘war on terror’ have highlighted.

These figures come from the Pavlov report; see A.I. Kokurin & N.V. Petrov, GULAG: Glavnoe upravlenie lagerei. 1918 – 1960 (Moscow, 2000), p. 433. The figure given in this source for the number of shootings in 1931 is obviously a misprint. The correct figure is given in the version of the Pavlov report printed in Artizov, Sigachev, Shevchuk & Khlopov, Reabilitatsiya: kak eto bylo, vol. 1 p. 76. According to an OGPU document published in Tragediya sovetskoi derevni, vol. 2 (Moscow, 2000), p. 809, in 1930 OGPU troikas sentenced 18,996 people to death, excluding East Siberia, Kazakhstan and Central Asia. This is the figure cited in Davies & Wheatcroft, The Years of Hunger…, p. 22.

In his telegram to Stalin of 10 February 1933 Eikhe pointed out that to accept the planned 100,000 deportees prior to the thawing of the rivers would require mobilising huge numbers of horses, which would severely disorganise forestry and agriculture. He also argued that even to accept the reduced number of deportees for West Siberia that he proposed would require massive preparatory efforts, one aspect of which would have to be building additional boats. For the text of the telegram see 1933 g. Nazinskaya tragediya (Tomsk, 2002), pp. 27 – 28. It is also printed in V. Danilov & S. Krasil'nikov (eds), Spetspereselentsy v Zapadnoi Sibiri 1933 – 1938 (Novosibirsk, 1994), p. 78. The latter also includes the related decision of the byuro of the kraikom of 9 February 1933.

Allen, Farm to Factory, chapter 4.

The usefulness—to the government—of the famine in bringing consumption and production into line was apparently pointed out by dr.Otto Schiller, the agricultural attaché of the German Embassy in Moscow, already in 1933; see J. Koshiw, ‘The 1932 – 33 Famine in the British Government Archives’, in W. Isajiw, Famine-Genocide in Ukraine, 1932 – 1933 (Toronto, 2003), p. 60.

V.P. Danilov & I.E. Zelenin, ‘Organizovannyi golod. K 70-letiyu obshchekrest'yanskoi tragedii’, Otechestvennaya istoriya, 2004, 5.

The idea that one of the motives of the party leadership was to punish the peasants for their poor work (or as the party leadership perceived it, sabotage) is also argued in Kondrashin & Penner, Golod …, pp. 209 – 210, 377 – 378.

The importance of policy is not in dispute. Davies & Wheatcroft, The Years of Hunger…, p. 434, write that ‘The fundamental cause of the deterioration in agriculture in 1928 – 33 was the unremitting state pressure on rural resources’. What is in dispute is the role of exogenous factors (two bad harvests) and of intent. Danilov & Zelenin differ from Davies & Wheatcroft on both these issues, and the present writer on just the latter.

Davies & Wheatcroft are quite right in thinking that two successive bad harvests, under the conditions prevailing in the Russian empire and in the USSR prior to 1948, normally led to famine. Furthermore, their arguments about the size of the 1931 and 1932 harvests are powerful and their estimates for these are the best currently available. However, it is necessary to consider not only why there was a famine but also why the number of victims was so large. Davies & Wheatcroft offer a better explanation of the first than of the second. After all, in 1945 and 1946 the harvest (according to official estimates) was only 47.3 and 39.6 million tons compared with 57 – 65 million tons and 55 – 60 million tons (the Davies & Wheatcroft estimates for 1931 and 1932). Nevertheless, the number of excess deaths in 1946 – 48 was only about 22% of that in 1931 – 34. In explaining the number of excess deaths in each case, one has to consider not only the size of the harvest but also the policies and intentions of the state. Even without two bad harvests, the policy of socialising the livestock of the Kazakhs without making adequate provision for their care might have generated a famine in Kazakhstan (although it would probably have led to fewer victims had the availability of grain been greater).

In The New English Bible, Hebrews, 12.11 reads ‘Discipline, no doubt, is never pleasant; at the time it seems painful, but in the end it yields for those who have been trained by it the peaceful harvest of an honest life’.

The need to discipline the collective farmers, and readiness to use tough measures to achieve this, were commonplace in the early 1930s. For example, in 1931 during the spring sowing campaign the use of floggings and beatings to raise discipline seems to have been widespread. This revival of the methods of serf-owners was especially used by activists of the ‘Unions for the struggle for discipline’. However, the local officials responsible seem subsequently to have been punished (Kondrashin & Penner, Golod…, p. 111).

There were some reductions. The initial 1932 grain procurement plan for the USSR of 20.56 million tons was reduced to 17.53 million tons by the end of November 1932; see Davies & Wheatcroft, The Years of Hunger…, pp. 181 – 185.

The party leadership did provide some famine relief; see Ibid., pp. 214 – 223, 424 – 426 and table 23.

Ibid., p. 440.

In addition to the economic arguments against this—the acute balance of payments crisis, its probable effect of crowding out machinery imports, and its likely negative effect on the USSR's credit rating—there was also a political objection. It would have undermined the USSR's image in the world. In August 1934 Molotov and Kaganovich suggested to Stalin importing 100,000 tons of grain from the Far East, while simultaneously exporting 100,000 tons of grain to Europe. This would have been a net import of zero, reduced the freight transport burden on the railways, and been financially profitable. Stalin, however, vetoed it on the ground that ‘The import of grain now, when abroad they are shouting about the shortage of grain in the USSR, might produce a political minus’ (Stalin i Kaganovich. Perepiska. 1931 – 1936 gg. (Moscow, 2001), pp. 461 – 462). A lot of attention has been given in recent years to how the activities of communists and fellow-travellers in the West (and in cases such as Gorky also in the USSR) supported Stalinist terror. A refusal to import grain out of apprehension about its political exploitation by anti-communists is an example of how the anti-communists (who could be relied on to exploit Soviet grain imports for propaganda purposes) also had a negative effect on the welfare of the Soviet population.

Davies & Wheatcroft, The Years of Hunger…, p. 238.

Penner, ‘Stalin and the Ital'ianka…’, pp. 40 – 48.

S. Kul'cyc'kyj, ‘Terror als methode. Der Hungergenozid in der Ukraine 1933’, Osteuropa, 54, 12, December 2004, p. 66.

Kosior's report was published in F.M. Rudych (ed.), Holod 1932 – 1933 rokiv na Ukraini (Kyiv, 1990), pp. 441 – 444; the crucial passages are on p. 443. This report seems not to be mentioned in Davies & Wheatcroft, The Years of Hunger….

Quoted from E.H. Carr, The New Society (London, 1960), p. 42. In 1933 in the Volga region there was a peasant rumour to the effect that the government used the famine in the way that some animal trainers used food deprivation. The aim of the latter was to break in the animals and train them to be obedient. The aim of the former was to discipline the peasants and make them obedient collective farmers (Kondrashin & Penner, Golod…, pp. 214 – 215). Although not the whole truth, this rumour may well have contained elements of the truth.

Extracts from the speeches of Stalin and Molotov at this meeting were published in V. Danilov, R. Manning & L. Viola (eds), Tragediya sovetskoi derevni, vol. 3 (Moscow, 2001), pp. 557 – 561. Stalin's speech is discussed in Davies & Wheatcroft, The Years of Hunger…, pp. 187 – 188. (Davies & Wheatcroft translate sokrushitel'nyi as ‘crushing’ rather than ‘knockout’.)

See for example the OGPU reports of 5 August 1932, 15 September 1932 and 22 September 1932 (Tragediya sovetskoi derevni, vol. 3, pp. 446 – 452, 472 – 476, and 488 – 489).

S. Krasil'nikov, Serp i molokh.Krest'yanskaya ssylka v Zapadnoi Sibiri v 1930-e gody (Moscow, 2003), p. 95; 1933 g. Nazinskaya tragediya, pp. 8 – 10. As far as the purpose of the 1933 deportations is concerned, O. Khlevniuk, The History of the Gulag (New Haven, 2004), pp. 55 – 56, has suggested that ‘It is possible that the leaders of the country supported the deportation of a large segment of the population in the European part of the USSR as a means both to reduce social tensions and to alleviate the acute food crisis’.

See 1933 g. Nazinskaya tragediya, pp. 8 – 10; Krasil'nikov, Serp i molokh…, p. 107. The 1933 deportations are not discussed in Davies & Wheatcroft, The Years of Hunger….

Krasil'nikov, Serp i molokh…, p. 106.

Khlevnyuk has suggested that the reason for the non-fulfillment of the 1933 deportation targets was the failure to organise efficiently the ‘special settlements’; see Khlevniuk, The History of the Gulag, p. 63.

For the text see Tragediya sovetskoi derevni, vol. 3, pp. 746 – 750.

Davies & Wheatcroft, The Years of Hunger…, p. 441 argue that the famine was ‘undesirable’ for the Soviet leadership. As a proposition about the external situation (the international image of the USSR), this is obviously correct. Also as a proposition about the internal situation, it is correct for the towns. The worker unrest and adverse effects on labour productivity were undesirable for the leadership. However, for the rural areas this characterisation ignores the famine's usefulness in eliminating ‘class enemies’ more efficiently than deportation. It also ignores the contribution to improving the grain balance made by reducing rural overpopulation. It also ignores the famine's contribution to disciplining/punishing/socialising/(re)educating the rural population.

I. Stalin, Voprosy Leninizma, 11th edn (Moscow, 1945), pp. 487 – 488, originally published in Pravda, 6 May 1935 (italics added). The corrected stenogram was published in V.A. Nevezhin (ed.), Zastol'nye rechi Stalina (Moscow and St Petersburg, 2003), pp. 84 – 91. For the uncorrected stenogram of a very similar speech Stalin made the previous day see Nevezhin, Zastol'nye rechi…, pp. 76 – 84.

E.H. Carr, What is History? (London, 1961), p. 75. Carr's remark refers explicitly to ‘collectivisation’, not the famine, but the ‘brutalities and abuses’ he refers to may well have included not just ‘dekulakisation’ but also the subsequent famine. They are often treated as part of the same process. There is also some temporal overlap.

Pravda, 28 September 1918.

B. Falk, Sowjetische Städte in der Hungersnot 1932/33 (Cologne, 2005), p. 31.

Ibid., table 16, p. 331.

For a study of the leadership's (not very successful) attempts to feed the workers during the famine see Ibid.

S.I. Golotik & V.V. Minaev, Naseleni i vlast’ (Moscow, 2004), p. 119, quoting two studies by Plotnikov about dekulakisation in the Urals.

Sovetskaya derevnya glazami …, vol. 3, book 1, pp. 774 – 787.

Tragediya sovetskoi derevni, vol. 3, p. 639. As Davies & Wheatcroft, The Years of Hunger…, p. 191, sensibly noted, ‘The desperate struggle of the state to exploit the peasants to the point of death was depicted as a righteous battle against counter – revolution’.

Rudych (ed.), Holod 1932 – 1933…, p. 443.

This perception was not confined to the leadership. N.A. Ivnitsky, Sud'ba raskulachennykh v SSSR (Moscow, 2004), p. 40, quotes a Komsomol member and teacher who witnessed ‘kulaks’ being packed in freight carriages prior to shipment to remote areas but did not feel any sympathy for them. As she wrote in her diary at the time, ‘Kulaks are kulaks, but people—they are people’. In other words, the kulaks as inhumane exploiters were not really people at all, even if they—and their wives and children—looked and sounded just like people. Hence they and their families did not deserve the kind of treatment appropriate to people.

O.V. Khlevnyuk et al. (eds), Stalin i Kaganovich. Perepiska. 1931 – 1936 gg. (Moscow, 2001), pp. 273 – 275.

For a well-informed discussion of the relationship between the famine and Soviet policies towards Ukraine see T. Martin, The Affirmative Action Empire (Ithaca, 2001), chapter 7; and T. Martin, ‘The 1932 – 33 Ukrainian Terror: New Documentation on Surveillance and the Thought Process of Stalin’, in W.Isajiw (ed.), Famine-Genocide in Ukraine, 1932 – 1933 (Toronto, 2003). For an overview of the role of the famine and the discussion of it in contemporary Ukraine see G.Kas'yanov, ‘Razrytaya mogila: golod 1932 – 1933 godov v Ukrainskoi istoriografii, politike i massovom soznanii’, Ab imperio, 2004, 3.

Davies & Wheatcroft, The Years of Hunger…, pp. 400 – 401. One famine which was clearly ‘intentional’ was the 1941 – 43 famine in occupied Kiev; see K.C. Berkhoff, Harvest of Despair (Cambridge, MA, 2004), chapter 7.

This is partially recognised in the Davies & Wheatcroft calculation of the number of famine victims. They include in the total of almost six million excess deaths from famine in 1930 – 33 approximately 300,000 (i.e. about 5%) in the OGPU system. These will have been mainly ‘kulaks’ and their family members who died during deportation or in prisons, ‘special settlements’ or camps. Naturally, the responsibility of the state for these deaths was greater than for those who died in their own villages or while in flight from them. Davies & Wheatcroft, The Years of Hunger…, also include an informative chapter (chapter 2) on the 1930 – 31 deportations.

In practice, the allocation of some of the deaths to one or another of these three groups might be somewhat arbitrary. Take the ‘kulaks’ and their family members who died en route to ‘special settlements’, in them, or while trying to escape from them. The policy of which they were victims was the ‘liquidation of the kulaks as a class’. This did not require them all to be killed, and their labour while in deportation was wanted, but their death helped the implementation of the policy.

• In March 1938, at the third of the big Moscow trials, Yagoda was accused of murdering Gorky. Applying the ‘accusation in a mirror’ logic, one could regard this as evidence that actually Stalin murdered Gorky. However, considering this matter in the light of this and additional evidence, Conquest concluded that the case against Stalin with respect to Gorky's death was non-proven (R. Conquest, The Great Terror: A Reassessment (London, 1990), p. 389). This indicates that traditionally the ‘accusation in a mirror’ type of argument has been regarded as only weak evidence. On its own, it is indeed only weak evidence. This article does not to rely on it alone, but strengthens it with respect to the 1931 – 34 famine by

  	1.	explaining its theoretical basis and contemporary use,
  	2.	giving a number of unambiguous examples where it is certainly correct to interpret Stalin's words using the ‘accusation in a mirror’ logic,
  	3.	presenting also several pieces of circumstantial evidence,
  	4.	explaining that the attitude of Bolsheviks to famines in general and to that of 1931 – 34 in particular was quite different from that of the general Western public today,
  	5.	pointing out that reliance on the disciplining effect of hunger does not reflect uniquely Bolshevik inhumanity but is inherent in certain types of industrialisation scenarios and was also part of the conventional wisdom of the British ruling class during the British industrial revolution,
  	6.	drawing attention to the need to explain both the fact of a famine in 1931 – 34 and the large number of victims of that famine, and
  	7.	qualifying the Davies & Wheatcroft argument that for the Soviet leadership the famine was ‘undesirable’ by pointing out that internally, in rural areas, it had a number of benefits for the Soviet leadership.

• The present author considers that these considerations taken together do indeed provide reasonable evidence that a starvation policy was one of the conjunctural factors contributing to the famine of 1931 – 34. In addition to the evidence presented in this article, there is also the argument from authority (the late V.P. Danilov).

Additional information

Notes on contributors

Michael Ellman

I am grateful for help, comments, discussion and criticism to K. Berkhoff, R. Binner, R.W. Davies, P. Ellman, M. Jansen, J. Keep, L. Viola and E. van Ree. The author alone is responsible for the interpretation and for the remaining errors.