Victims of the Soviet Penal System in the Pre-war Years: A First Approach on the Basis of Archival Evidence, By: J. ARCH GETTY, GABOR T. RITTERSPORN, VIKTOR N. ZEMSKOV et.al.

By: J. ARCH GETTY, GABOR T. RITTERSPORN, VIKTOR N. ZEMSKOV et.al.

Source: American Historical Review  http://web.archive.org/web/20080611064213/http:/www.etext.org/Politics/Staljin/Staljin/articles/AHR/AHR.html  

The Great purges of the 1930s were a maelstrom of political violence that engulfed all levels of society and all walks of life. Often thought to have begun in 1934 with the assassination of Politburo member Sergei Kirov, the repression first struck former political dissidents in 1935-1936. It then widened and reached its apogee in 1937-1938 with the arrest and imprisonment or execution of a large proportion of the Communist Party Central Committee, the military high command, and the state bureaucracy. Eventually, millions of ordinary Soviet citizens were drawn into the expanding terror.

Debate in the West about the precise numbers of victims has appeared in the scholarly press for several years and has been characterized by wide disparity, often of several millions, between high and low estimates. Using census and other data, scholars have put forward conflicting computations of birth, mortality, and arrests in order to calculate levels of famine deaths due to agricultural collectivization (1932-1933), victims of the Great Terror (1936-1939), and total “unnatural” population loss in the Stalin period. Anton Antonov-Ovseenko, Robert Conquest, Steven Rosefielde, and others have posited relatively high estimates (see Table 1). On the other hand, Stephen Wheatcroft and others working from the same sources have put forth lower totals. Both “high” and “low” estimators have bemoaned the lack of solid archival evidence and have claimed that should such materials become available, they would confirm the author’s projection. The debate, along with disputes on the “totalitarian” nature of the Stalinist regime, the importance of Joseph Stalin’s personality, and the place of social history in Soviet studies, has polarized the field into two main camps, perhaps unfortunately labeled “Cold Warriors” and “revisionists.” Revisionists have accused the other side of using second- hand sources and presenting figures that are impossible to justify, while the proponents of high estimates have criticized revisionists for refusing to accept grisly facts and even for defending Stalin. Both sides have accused the other of sloppy or incompetent scholarship.

Now, for the first time, Soviet secret police documents are available that permit us to narrow sharply the range of estimates of victims of the Great Purges. These materials are from the archival records of the Secretariat of GULAG, the Main Camp Administration of the NKVD/MVD (the USSR Ministry of the Interior). They were housed in the formerly “special” (that is, closed) sections of the Central State Archive of the October Revolution of the USSR (TsGAOR), which is now part of the newly organized State Archive of the Russian Federation (GARF). A few Moscow scholars (among them V. N. Zemskov) had access to some of them in the past but were not allowed to cite them properly. Now, according to the liberalized access regulations in Russian archives, scholars are able to consult these documents and to publish exact citations. (See “A Note on Sources” at the end of this article.)

We propose to deal here only with quantitative elements of the terror, with what we can now document of the scale of the repression. Of course, such a cold numerical approach risks overshadowing the individual personal and psychological horror of the event. Millions of lives were unjustly taken or destroyed in the Stalin period; the scale of suffering is almost impossible to comprehend. The horrifying irrationality of the carnage involves no debatable moral questions - destruction of people can have no pros and cons. There has been a tendency to accuse “low estimators” of somehow justifying or defending Stalin (as if the deaths of 3 million famine victims were somehow less blameworthy than 7 million).

Scholars and commentators will make use of the data as they choose, and it is not likely that this new information will end the debates. Still, it seems a useful step to present the first available archival evidence on the scale of the Great Terror. Admittedly, our figures are far from being complete and sometimes pose almost as many questions as they answer. They nevertheless give a fairly accurate picture of the orders of magnitude involved and show the possibilities and limits of the data presently available.

The penal system admrnistered by the NKVD (Peoples' Commissariat of Internal Affairs) in the 1930s had several components: prisons, labor camps, and labor colonies, as well as "special settlements" and various types of non-custodial supervision. Generally speaking, the first stop for an arrested person was a prison, where an investigation and interrogation led to conviction or, more rarely, release. After sentencing, most victims were sent to: one of the labor camps or colonies to serve their terms. In December 1940, the jails of the USSR had a theoretical prescribed capacity of 234,000, although they then held twice that number. Considering this-and comparing the levels of prison populations given in the Appendixes for the 1930s and 1940s one can assume that the size of the prison system was probably not much different in the 1930s.

Second, we find a system of labor camps. These were the terrible “hard regime” camps populated by dangerous common criminals, those important politicals” the regime consigned to severe punishment, and, as a rule, by other people sentenced to more than three years of detention. On March 1, 1940, at the end of the Great Purges, there were 53 corrective labor camps (ispravitel’no-trudovye lageri: ITL) of the GULAG system holding some 1.3 million inmates. Most of the data cited in this article bear on the GULAG camps, some of which had a multitude of subdivisions spreading over vast territories and holding large numbers of people. BAMLAG, the largest camp in the period under review, held more than 260,000 inmates at the beginning of 1939, and SEVVOSTLAG (the notorious Kolyma complex) some 138,000.

Third came a network of 425 “corrective labor colonies” of varying types. These colonies were meant to confine prisoners serving short sentences, but this rule varied with time. The majority of these colonies were organized to produce for the economy and housed some 315,000 persons in 1940. They were nevertheless under the control of the NKVD and were managed-like the rest of the colony network-by its regional administrations. Additionally, there were 90 children’s homes under the auspices of the NKVD.

Fourth, there was the network of “special resettlements.” In the 1930s, these areas were populated largely by peasant families deported from the central districts as “kulaks” (well-to-do peasants) during the forced collectivization of the early 1930s. Few victims of the Great Purges of 1936-1939 were so exiled or put under other forms of non-custodial supervision: in 1937-1938, only 2.1 percent of all those sentenced on charges investigated by the political police fell into this category. This is why we will not treat exile extensively below.

Finally, there was a system of non-custodial “corrective work” (ispravitel’no-trudovye raboty), which included various penalties and fines. These were quite

common throughout the 1930s-they constituted 48 percent of all court sentences in 1935-and the numbers of such convictions grew under the several laws on labor discipline passed on the eve of the war. Typically, such offenders were condemned to up to one year at “corrective labor,” the penalty consisting of work at the usual place of one’s employment, with up to 25 percent reduction of wage and loss of credit for this work toward the length of service that gave the right to social benefits (specific allocations, vacation, pension). More than 1.7 million persons received such a sentence in the course of 1940 and almost all of them worked in their usual jobs “without deprivation of freedom.” As with resettlements, this correctional system largely falls outside the scope of the Great Terror.

Figure A provides the annual totals for the detained population (GULAG camps, labor colonies, and “kulak” resettlements, minus prisons) in the years of the Great Purges. It shows that, despite previously accepted-and fairly inflated-figures to the contrary, the total camp and exile population does not seem to have exceeded 3.5 million before the war. Were we to extrapolate from the fragmentary prison data we do have (see the Appendixe’s), we might reasonably add a figure of 300,000-500,000 for each year, to put the maximum total detained population at around 3 million in the period of the Great Purges.

Figure A: Camp, Colony, and "Kulak" Exile Populations, USSR, 1935-1940

Mainstream published estimates of the total numbers of “victims of repression” in the late 1930s have ranged from Dmitrii Volkogonov's 3.5 million to Ol'ga Shatunovskaia's nearly 20 million. (See Table 1.) The bases for these assessments are unclear in most cases and seem to have come from guesses, rumors, or extrapolations from isolated local observations. As the table shows, the documentable numbers of victims are much smaller.

We now have archival data from the police and judiciary on several categories of repression in several periods: arrests, prison and camp growth, and executions in 1937-1938, and deaths in custody in the 1930s and the Stalin period generally. Runs of data on arrests, charges, sentences, and custodial populations in the 1930s unfortunately reflect the simultaneous actions of several punitive agencies including the secret police, procuracy, courts, and others, each of which kept their own records according to their own statistical needs. No single agency (not even the secret police) kept a “master list” reflecting the totality of repression. Great care is therefore needed to untangle the disparate events and actors in the penal process.

Table 1 Current Estimates of the Scale of Stalinist Repression

	1937-38 total arrests	1938 camp population	1938 prison and camp population	1952 camp population	1937-38 camp deaths	1937-38 executions	1921-53 executions
Anton Antonov- Ovsenko	18.8 million		16 million				7 million
Roy A. Medvedev	5-7 million					0.5-0.6 million
Ol’ga Shatunovskaia	19.8 million					7 million
Dmitri Volkogonov	3.5-4.5 million
Robert Conquest	7-8 million	~7 million	~8 million	12 million	2 million	1 million
Documentable	~2.5 million	1.9 million	2.0 million	2.5 million	160,084	681,692	799,455

A 1953 statistical report on cases initiated or investigated by the NKVD provides data on arrests and on the purported reasons for them. According to these figures, 1,575,259 people were arrested by the security police in the course of 1937-1938, 87.1 percent of them on political grounds. Some 1,344,923, or 85.4 percent, of the people the secret police arrested in 1937-1938 were convicted. To be sure, the 1,575,259 people in the 1953 report do not comprise the total of 1937-1938 arrests. Court statistics put the number of prosecutions for infractions unrelated to “counterrevolutionary” charges at 1,566,185, but it is unlikely that all persons in this cohort count in the arrest figures. Especially if their sentence was non-custodial, such persons were often not formally arrested. After all, 53.1 percent of all court decisions involved non-custodial sentences in 1937 and 58.7 percent in 1938, and the sum total of those who were executed or incarcerated yields 647,438 persons in categories other than “counterrevolution.” Even if we remember that during the Great Purges the authorities were by far more inclined to detain suspects than in other times, it seems difficult to arrive at an estimate as high as 2.5 million arrests on all charges in 1937-1938.

Although we do not have exact figures for arrests in 1937-1938, we do know that the population of the camps increased by 175,487 in 1937. and 320,828 in 1938 (it had declined in 1936). The population of all labor camps, labor colonies, and prisons on January 1, 1939, near the end of the Great Purges; was 2,022,976 persons. This gives us a total increase in the custodial population in 1937-1938 of 1,006,030. Nevertheless, we must add to these data the number of those who had been arrested but not sent to camps, either because they were part of a small contingent released sometime later or because they were executed.

As Table 1 shows, popular estimates of executions in the Great Purges of 1937-1938 vary from 500,000 to 7 million. We do not have exact figures for the numbers of executions in these years, but we can now narrow the range considerably. We know that between October 1, 1936, and September 30, 1938, the Military Board of the Supreme Court, sitting in 60 cities and towns, sentenced 30,514 persons to be shot. According to a press release of the KGB, 786,098 persons were sentenced to death “for counterrevolutionary and state crimes” by various courts and extra-judicial bodies between 1930 and 1953. It seems that 681,692 people, or 86.7 percent of the number for this 23-year-period were shot in 1937-1938 (compared to 1,118 persons in 1936). A certain number of these unfortunates had been arrested before 1937, including exiled and imprisoned ex-oppositionists who were summarily killed in the autumn of 1937. More important, however, our figures on 1937-1938 executions are not entirely comparable to those quoted in the press release. Coming from a 1953 statistical report “on the quantity of people convicted on cases of NKVD bodies,” they also refer to victims who had not been arrested for political reasons, whereas the communique concerns only persons persecuted for “counterrevolutionary offenses.” In any event, the data available at this point make it clear that the number shot in the two worst purge years was more likely a question of hundreds of thousands than of millions.

Of course, aside from executions in the terror of 1937-1938, many others died in the regime’s custody in the decade of the 1930s. If we add the figure we have for executions up to 1940 to the number of persons who died in GULAG camps and the few figures we have found so far on mortality in prisons and labor colonies, then add to this the number of peasants known to have died in exile, we reach the figure of 1,473,424. To be sure, of 1,802,392 alleged kulaks and their relatives who had been banished in 1930-1931, only 1,317,022 were still living at their places of exile by January 1, 1932. (Many people escaped: their number is given as 207,010 only for the year of 1932.) But even if we put at hundreds of thousands the casualties of the most chaotic period of collectivization (deaths in exile, rather than from starvation in the 1932 famine), plus later victims of different categories for which we have no data, it is unlikely that “custodial mortality” figures of the 1930s would reach 2 million: a huge number of “excess deaths” but far below most prevailing estimates. Although the figures we can document for deaths related to Soviet penal policy are rough and inexact, the available sources provide a reliable order of magnitude, at least for the pre-war period.

Turning to executions and custodial deaths in the entire Stalin period, we know that, between 1934 and 1953, 1,053,829 persons died in the camps of the GULAG. We have data to the effect that some 86,582 people perished in prisons between 1939 and 1951. (We do not yet know exactly how many died in labor colonies.) We also know that, between 1930 and 1952-1953, 786,098 “counter-revolutionaries” were executed (or, according to another source, more than 775,866 persons “on cases of the police” and for “political crimes”). Finally, we know that, from 1932 through 1940, 389,521 peasants died in places of “kulak” resettlement. Adding these figures together would produce a total of a little more than 2.3 million, but this can in no way be taken as an exact number. First of all, there is a possible overlap between the numbers given for GULAG camp deaths and “political” executions as well as between the latter and other victims of the 1937-1938 mass purges and perhaps also other categories falling under police jurisdiction. Double-counting would deflate the 2.3 million figure. On the other hand, the 2.3 million does not include several suspected categories of death in custody. It does not include, for example, deaths among deportees during and after the war as well as among categories of exiles other than “kulaks.” Still, we have some reason to believe that the new numbers for GULAG and prison deaths, executions as well as deaths in peasant exile, are likely to bring us within a much narrower range of error than the estimates proposed by the majority of authors who have written on the subject.

Table 2. Age and Gender Structure of GULAG Population (as of January 1 of each year)

AGE/SEX	Percent of GULAG Population			Percent of USSR Population
	1934	1937	1940	January 1937	January 1939
up to 18 years of age	1.2	0.7	0.5	5.0	-
19-24	23.8	12.0	9.6	10.3	-
25-30	26.2	47.0	34.8	11.7	33.0
31-40	28.1	26.3	30.0	13.8	-
41-50	16.0	10.7	16.7	8.7	9.0
50+	4.7	3.3	8.4	11.9	13.0
Women	5.9	6.1	8.1	52.7	-

We now have some information about the demographic composition of the GULAG's prisoners. In terms of gender, there are few surprises. As Table 2 shows, women constituted a minority of hard regime camp inmates, although their share reached almost 13 percent by 1943 and 24 percent by 1945. They accounted for no more than 11 percent of the people prosecuted by the court system until the late 1930s, then the demographic situation of the war years increased their part to more than 40 percent by 1944; and, even though this proportion diminished afterward, it did not descend below 20 percent until 1955.

As we look at Table 2, the prominence of persons between 25 and 40 years of age among labor camp inmates is not surprising. A shift can be observed between 1934 and 1940. The generation that grew up in the tumult of war, civil war and revolution and came of age in the New Economic Policy era continued to constitute a cohort more exposed to penal sanctions than the rest of society. Thus people between ages 19 and 24 in 1934 are likely to account for the large over-representation of the age group 25 to 30 in 1937 and of the 31 to 35 cohort on the eve of the war. Those in the 51 to 60 and especially 41 to 50 age ranges, however, seem to be most vulnerable to repression in the wake of crises like collectivization and the Great Purges. The presence of persons between ages 18 and 21 also becomes notable in the camps by March 1940, when they made up 9.3 percent of the inmates (their share in the 1937 population was 6.4 percerit).

In fact, it gives one pause to reflect that 1.2 percent of strict regime camp detainees were 18 or younger in 1934 and that, by 1941, their share nearly reached the proportion of those between 16 and 18 in the country’s population. From mid-1935 to the beginning of 1940,155,506 juveniles between the ages of 12 and 18 passed through the labor colonies. Some 68,927 of them had been convicted of a crime and 86,579 had not. The large proportion of unconvicted young detainees indicates that they were likely to be incarcerated by extra-judicial bodies, as was a high proportion of adult inmates not sentenced by courts between 1938 and 1940. Nevertheless, political reasons did not play a predominant role in the conviction of minors. The ordeal of collectivization and the ensuing famine as well as the turmoil of mass migration from countryside to cities dramatically increased the number of orphans, abandoned children, and single-parent house-holds and weakened the family as well as the social integration of some categories of youth. Juvenile delinquency became a serious concern for the authorities by the spring of 1935, when they ordered that the courts were entitled to apply “all penal sanctions” to children having reached 12 years and guilty of “theft, violence, bodily harm, mutilation, murder and attempted murder.”

Records show that 10,413 youngsters between 12 and 16 years of age were sentenced by the courts of the Russian Federation in the second half of 1935 and the first half of 1936; 77.7 percent of them were accused of theft (as opposed to 43.8 percent of those in the 16 to 18 group) and 7.1 percent of violent crimes. At this time, when the overall proportion of custo,dial sentences did not exceed 44 percent in the republic, 63.5 percent of the youngest offenders (and 59.4 percent between 16 and 18) were sent to detention. In addition, there was a tendency to apply the 1935 decree to infractions it did not cover; thus, despite instructions to the contrary, 43 juveniles were sentenced for alleged misconduct in office [!] by mid-1936 and 36 youngsters under 16 were so sentenced between 1937 and 1939. The sources show, incidentally, that the procuracy suggested that people below 18 years of age should not be confined in ordinary places of detention, and there is reason to believe that it also vainly protested against a directive of the camp administration stipulating that “the stay of minors in labor colonies is not limited by the terms of court sentences.”

Table 3. Data on 10,366 Juvenile Camp Inmates, April 1, 1939

Sentenced for:	No.	Percent of All Sentences	Adults: Percent of All Sentences January 1, 1939
“Counterrevolutionary offenses”	160	1.6	34.5
Dangerous crimes against the administrative order	929	9.0	14.8
Banditry	97	0.9	1.4
Misconduct in office	60	0.6	6.1
Crimes against persons	434	4.2	4.8
Crimes against property	2,507	24.4	12.1
Teft of public property	22	0.2	2.1
Being “socially harmful and dangerous elements”	5,838	56.9	21.7
Violating the law on internal passports	115	1.1	2.1
Other crimes	204

At any rate, 24,700 children and adolescents up to 16 years of age appeared in courts in 1938 and 33,000 in the course of the following year, an increase that reflects a hardening penal practice. Table 3 indicates, however, that even if juveniles could be detained for political reasops, this motive did not account for a high proportion of the youngest camp inmates, even in the wake of the Great Purges. Although these data denote a tendency to imprisonjuveniles almost in the same proportions as adults if they were accused of the most serious crimes, they also show the penal system's proclivity to impose custodial sentences on youngsters more readily than on grown-ups.

Table 4 shows the national origin of the majority of labor camp inmates on January 1, 1937-1940, alongside the ethnic composition of the USSR according to the working materials of the (suppressed) 1937 and (published) 1939 censuses. In comparison with their weight in the general population, Russians, Belorussians, Turkmen, Germans, and Poles were over-represented in the camps by 1939; Germans and Poles being especially hard-hit. On the other hand, Ukrainians, Jews, Central Asians (except Turkmen) and people from the Caucasus were less represented in the GULAG system than in the population of the country; as national groups, they suffered proportionately less in the 1937-1938 terror.

Table 4. Ethnic Groups in GULAG Camps, January 1, 1937-1940

Ethnic group	1937	1938	1939	1940	1937 camps %	1937 census %	1939 camps %	1939 census %	over (+) under (-) representation (camps and census)
									1937	1939
Russians	494,827	621,733	830,491	820,089	60.28	58.07	63.05	58.09	+2.21	+4.96
Ukrainians	138,318	141,447	181,905	196,283	16.85	16.33	13.81	16.47	+0.52	-2.66
Belorussians	39,238	49,818	44,785	49,743	4.78	3.01	3.40	3.09	+1.57	+0.31
Tattars	-	22,916	24,894	28,232	-	1.35	1.89	2.52	-	-0.63
Uzbeks	29,141	19,927	24,499	26,888	3.55	2.81	1.86	2.84	+0.74	-0.98
Jews	11,903	12,953	19,758	21,510	1.45	1.65	1.50	1.77	-0.20	-0.27
Germans	-	998	18,572	18,822	-	0.71	1.41	0.84	-	+0.57
Kazakhs	-	11,956	17,123	20,166	-	1.77	1.30	1.82	-	-0.52
Poles	-	6,975	16,860	16,133	-	0.39	1.28	0.37	-	+0.91
Georgians	4,351	6,974	11,723	12,099	0.53	1.24	0.89	1.32	-0.71	-0.43
Armenians	5,089	6,975	11,064	10,755	0.62	1.22	0.84	1.26	-0.60	-0.42
Turkmen	-	4,982	9,352	9,411	-	0.46	0.71	0.46	-	+0.23
Latvian	-	1,191	4,742	5,400	-	0.04	0.58	0.07	-	+0.51
Finns	-	997	2,371	2,750	-	0.09	0.29	0.08	-	+0.21

If ethnic groups for whom camp figures are unavailable in 1937 were too weakly represented to be counted, then Table 4 accurately demonstrates the statistical impact of the terror on different nationalities. Because we know that the party/state administration was heavily staffed by Russians and that many members of the party elite and economic leadership were of Polish and German background, the changes in the ethnic composition seem to indicate a terror aimed more at the elite than at particular national groups per se. To be sure, a sizable proportion of citizens of Polish and German origin living in border areas suffered several waves of “cleansing” for their alleged unreliability. In addition, wherever they resided, they were likely to be accused of political sympathies with states with which relations were strained, especially at a time when the authorities suspected fifth columns throughout the country and ordered a clampdown on “spies and nationalists.” This circumstance must have contributed to the fact that, in early 1939, when GULAG inmates made up 0.77 percent of the country’s pppulation, some 2.7 percent and 1.3 percent of these ethnic groups were in hard regime camps, as well as about 1.3 percent of all Koreans, 1.7 percent of all Estonians, 1.9 percent of all Finns, and 3.2 percent of all Lithuanians, compared to approximately 0.85 percent of all Belorussians, 0.84 percent of all Russians, 0.65 percent of all Ukrainians, and 0.61 percent of all Jews. The national group suffering the most in proportional terms was the Latvians, who were heavily represented in the party and state administration and of whose total census population a staggering 3.7 percent was in strict regime camps alone. The hypothesis of an increasingly anti-elite orientation of the penal policy is supported by data on the educational levels of labor camp inmates. Table 5 shows the educational background of hard regime camp inmates on January 1, 1937, alongside educational levels for the population as a whole in 1937. Even allowing for the rise in educatiorial levels in the general population between 1937 and 1940, it seems clear that the purge hit those with higher educational levels more severely. Although less educated common folk heavily outnumbered the “intelligentsia” in the camps, those who had studied in institutions of higher or secondary education were proportionally nearly twice as numerous in the GULAG system as they were in society at large, while those with elementary (or no) education were under-represented.

Table 5. Educational Levels of the GULAG Population versus the USSR as a Whole, 1937

School Achievemen:	GULAG Population 1937 (%)	USSR Population 1937 (%)
higher	1.0	0.6
secondary	8.9	4.3
elementary	49.3	38.3
semi-literate	32.4	-
illiterate	8.4	39.0

Moreover, in the years spanning the Great Terror, the proportion of the camp population with some education rose significantly, while that of less educated people declined. From 1934 to 1941, the segment of the camp population with higher education tripled and the proportion with secondary education doubled. Again, however, care must be used in interpreting these data, because educational levels in the population as a whole were increasing steadily during the decade of the 1930s. We lack detailed annual education data for the period and especially statistics on the share of people with college and high school instruction in the population of the late 1930s and early l940s. Thus it would be dangerous to draw firm conclusions, even though the available evidence strongly suggests that the terror intensified against the educated elite. It comprised 12.8 percent of the population of hard regime camps by 1941, compared to 6.3 percent in 1934. As Table 6 indicates, the number of detainees with higher and secondary education grew much faster than the rest of the GULAG population.

It is commonly believed that most of the prisoners of the “Gulag Archipelago” had been arrested and sentenced for political offenses falling under one of the headings of “counterrevolutionary offenses” (Article 58 in the criminal code). It is also common wisdom that many people arrested for other reasons were accused of political crimes for propaganda value. The available evidence does not bear out this view, but it does suggest considerabte ambiguity in definitions of “political crimes.” Table 7 shows the breakdown of labor camp inmates for selected years, according to the offense for which they were sentenced. Although the presence of alleged counterrevolutionaries is impressive, it turns out that ostensibly non-political detainees heavily outnumbered “politicals.”

Table 6. Percentage of Increase in Detainees by Educational background in GULAG Camps

Education:	1934-1936	1936-1939	1939-1941
higher	+47.5	+69.6	+25.6
secondary	+54.1	+48.0	+23.5
elementary and less	+37.9	+34.4	+7.9

In view of the murderous campaign of 1932-1933 against pilferers of state and collective farm property. and of the fact that in 1951 the number of prisoners convicted for this offense largely outstripped that of all categories of “counterrevolutionaries,” their share seems at first glance suspiciously low in Table 7, especially in 1940. One explanation for the relatively low proportion of inmates convicted under the “Law of August 7, 1932”-which had prescribed the death penalty or ten years of hard labor for theft of state property-is an unpublished decree of January 1936 ordering the review of the cases of all inmates convicted under the terms of this Draconian law before 1935. The overwhelming majority of these people had been condemned between 1932 and 1934, and four-fifths of this cohort saw their sentences reduced by August 1936 (including 40,789 people who were immediately released). Another possible explanation is that many people benefited from a directive reorienting the drive against major offenders and from reviews of their convictions that led by the end of 1933 to modifications of 50 percent of the verdicts from the previous seventeen months. This state of affairs seems to account for the considerable confusion in the records concerning the implementation of the “Law of August 7” and for the fact that, while claiming that the number of persons sentenced under its terms was between 100,000 and 180,000, officials were reluctant to advance exact figures even as late as the spring of l936.

Table 7. Offenses of GULAG Population (by percent as of January 1 of each year)

Sentenced for:	1934	1936	1940
“Counterrevolutionary offenses”	26.5	12.6	33.1
Dangerous crimes against the administrative order, including	15.2	17.7	3.6
-banditry	3.9	3.2	2.4
Other crimes against the administrative order, including:	1.3	-	13.9
-specualtion	1.3	1.1	2.4
-“hooliganism”	-	-	7.3
Misconduct in office, Economic crimes	7.5	10.6	7.3
Crimes against persons	4.7	5.5	5.2
Crimes against property	15.9	22.3	12.1
Teft of public property	18.3	14.2	1.9
“Socially harmful and dangerous elements”	8.0	11.5	18.9
Violation of the law on internal passports	-	2.3	1.3
Military offenses	0.6	0.8	0.7
Other delicts	2.0	2.6	3.3

The category of “socially harmful and dangerous elements” and the manner it was put to use must also warn us not to accept the definitions of “counter-revolutionaries” in our sources. Article 7 of the penal code stated that “to persons having committed socially dangerous acts or representing danger through their relation(s) with the criminal milieu or through their past activities, measures of social defense of a judicial-corrective, medical or medico-pedagogical character are applied.” Nevertheless, it failed to specify penalties except to indicate in Article 35 that these persons could be subjected to internal exile, without giving the slightest hint of the sentences courts were entitled to pass. The definition of the offense and the corresponding penalty were more than vague, but this did not prevent extra-judicial bodies of the secret police from singling out “harmful” and “dangerous” people among “recidivists [and] persons associated with the criminal milieu conducting a parasitic way of life etc.” This information comes from an appeal Co the top leadership by the procurator general, who was proposing to restrict the sentencing powers of the NKVD Special Board at the beginning of 1936 but not insofar as “dangerous elements” were concerned.

Although the procurator of the USSR, Andrei Vyshinskii, valued procedural precision, his office does not appear to have objected to the launching in August 1937 of a lethal “mass operation” targeting “criminals (bandits, robbers, recidivist thieves, professional smugglers, recidivist swindlers, cattle thieves) engaged in criminal activities and associated with the criminal milieu)”-whether or not they were actually guilty of any specific offense at the moment-and connecting these common criminals to a wide range of supposedly “anti-Soviet” and “counter-revolutionary” groups, from “kulaks” to former members of forbidden political parties, former oppositionists, and alleged terrorists. Clearly, the regime saw a political threat in the conduct, and indeed in the sheer existence, of “dangerous” persons. The secret directive of 1937 was no dead letter: the records suggest that it led to the arrest of a great number of people. some of whom were hardly more than notorious hooligans and yet were sometimes sent to the firing squad.

Some 103,513 “socially harmful and dangerous elements” were held in hard regime camps as of January 1937, and the numbec grew to 285,831 in early 1939, when, as Table 3 shows, they made up a record 21.7 percent of all detainees (and 56.9 percent ofjuvenile detainees). But the proportion (and also the number) of “dangerous” persons began to decline by January 1940 and that of “hooligans” started to rise, until the size of their contingent came close to that of the “harmful elements” by 1941, in part because of toughened legislation concerning rowdies. A total of 108,357 persons were sentenced in 1939 for “hooliganism”; in the course of the next year, 199,813 convicts fell into this category. But by 1948, the proportion of “hooligans” among camp inmates was 2.1 percent, whereas that of "dangerous elements” fell to 0.1 percent. No doubt the same offense in the 1930s could be regarded as “socially dangerous” and in the 1940s as “hooliganism.”

“Socially harmful” people may have been victims of political repression, but it would be far-fetched to presume that the unjust punishment they received was a response to conscious acts of opposition to the regime. Having observed this, we must remember that the great majority of those sentenced for “counter-revolutionary offenses” had never committed any act deliberately directed against the Soviet system and even continued to remain faithful to the Bolshevik cause, notwithstanding their victimization. From this point of view, the regime’s distinction between “political” and “non-political” offenders is of doubtful relevance. Unless we are prepared to accept broad Stalinist definitions of “counterrevolutionary” offenses or the equally tendentious Western categorization of all arrests during Stalin’s time (even those for crimes punishable in any society) as political, we should devise ways to separate ordinary criminality from genuine opposition to the system as well as from other reasons for which people were subjected to penal repression.

At any rate, the Appendix figures show that from 1934 to 1953, a minority of the labor camp inmates had been formallv convicted of “counterrevolutionary crimes.” Our data on sentencing policy are incomplete for the period before 1937, but they permit us to advance some estimates of orders of magnitude. Thus we can calculate that only about 11 percent of the more than 5.3 million persons sentenced by courts and extrajudicial bodies between 1933 and 1935 represented “cases of the OGPU/NKVD” of which. as we have seen, a relatively high proportion had not been considered “political.” Some 28 percent of the almost 5 million people convicted by various courts and NKVD boards in 1937-1939 were sentenced “from cases of the security police,” mostly under the pretext of “counterrevolutionary offenses.” But while the judiciary and the Special Board of the NKVD/MVD subjected nearly 31 million persons to penalties in the period 1940-1952, only 4.8 percent (though a sizable 1.5 million persons) fell under Article 58. By contrast, more than twice as many (11 percent) of all people sentenced in these years were charged with appropriating public property.

Table 8. GULAG Population according to Sentencing Authority (Percentages as of January 1)

Jurisdiction	1934	1935	1936	1937	1938	1939	1940	1941
Police bodies, including:	42.2	41.3	33.7	30.9	49.8	59.4	54.5	38.7
-the Special Board of the NKVD					3.7	8.3	9.4	8.2
-the “Special Troikas” of 1937-1938						23.3	25.4	17.2
Courts and Tribunals	57.8	58.7	66.3	69.1	50.2	40.6	45.5	58.6

It turns out that by far the largest group of those sentenced between 1940 and 1952 consisted of people accused of violating laws devised to strengthen labor discipline, ranging from unauthorized absence from work to dodging mobilization for work in agriculture, to failing to meet the compulsory minimum of work in the collective farm. Although the judiciary jargon called them “wartime decrees,” most of them remained in force until 1956. More than 17 million people had been convicted under their terms between 1940 and 1952 (albeit “only” 3.9 million of them were sentenced to detention), comprising half (55.3 percent) of all the period’s sentences. One may wonder if acts infringing on proprietary prerogatives and labor relations in a state that is virtually the only proprietor and practically the only employer do not bear some relation to politics. But if we leave aside this dilemma as well as the year 1936, for which our data are too fragmentary, we can conclude that, on the whole, only about 8.4 percent of the sentences of courts and extra-judicial bodies were rendered “on cases of the secret police” and for alleged political reasons between 1933 and 1953.

From 1934, when many believe the terror was mounting, to 1937-1938, the camp proportion of “counterrevolutionaries” actually declined. Table 8 shows that so did the proportion in the strict regime camp population of those who had been sent there by specific police bodies.

Even though the number of people convicted “on cases of the NKVD” more than tripled from 1934 to 1935, a careful look at the sources shows that many sentences had hardly anything to do with “political” cases. Data on the arrested “counterrevolutionaries” show a 17 percent growth due to an increase in the number of people accused of “anti-Soviet agitation” by a factor of 2.6. As for sentences in 1935, 44.6 percent of them were rendered by regional NKVD “troikas” (tribunals), which did not deal with “political” affairs. Another 43 percent were passed by regular courts, but fewer than 35,000 of the more than 118,000 people concerned had been “counterrevolutionaries.” To be sure, the quantity of “political” sentences increased, compared to the previous year. In 1936, however, the NKVD arrested the same number of “counterrevolutionaries” as in 1934, which does not seem to show steadily intensifying political repression. Similarly, the continually decreasing number of people shot in cases initiated by the secret police and the constantly diminishing share (as well as aggregate number) of “counterrevolutionaries” in hard regime camps between 1934 and 1937 casts doubt on the idea of “mounting” repression in this period.

The abolition of the OGPU, a degree of uncertainty concerning the sentencing privileges of the new NKVD, and attempts to transfer the bulk of “political” cases to the jurisdiction of military tribunals as well as to the special boards of regional courts and the Supreme Court suggest that the penal policy of more or less ordinary judicial instances, whose statistics are available, is indicative of the general trend of 1935-1936. The data are unfortunately incomplete, but we have information on at least 30,174 “counterrevolutionaries” who were sentenced by civilian and military courts in 1935, in the wake of the Kirov assassination, and on 19,080 people who were prosecuted by the same courts for supposedly political offenses in the first half of the next year. Most of this growth is attributable to the increased frequency of “anti-Soviet agitation,” which accounted for 46.8 percent of the cases before the courts of the Russian Federation in the first six months of 1935, and 71.9 percent in the corresponding period of the next year. The loose application of this charge did not always sit well in high places, and the people’s commissar of justice along with the prosecutor general warned top decision-makers of the consequences of an excessive use of the more than vague legislation on “counterrevolutionary agitation.” The prosecutor general had a heated exchange of letters with the head of the security police that raised the possibility of limiting NKVD jurisdiction in this matter.

There was a tendency to diminish rather than inflate the share of “political” cases in 1936. Even the chairman of the ominous Military Collegium of the Supreme Court noted in December 1936 that the numbet of “counterrevolutionaries” convicted by his bench and its subordinate courts in the first nine months of the year was 34.4 percent less than in the same period of 1935. The number of prosecutions had grown only for two categories of crimes. Characteristically enough, these were espionage and sabotage, and their frequency increased, especially in the third quarter of 1936.

It is from that time, late 1936, and not from late 1934 that the number of “counterrevolutionaries” (as well as the cohort sentenced by the NKVD) began to swell dramatically, above all in the wake of the launching of wholesale “mass operations” during the summer of 1937 that victimized “socially harmful” people alongside a wide range of purported political delinquents. The documents that ordered the mass “repression of former kulaks, criminals, and anti-Soviet elements” through decisions of newly organized “Special Troikas” of the secret police specified that the operation had to be completed within four months and even set “control figures” for the numbers of people to be shot and imprisoned. The relevant instruction foresaw 72,950 executions and 186,500 new detainees as the outcome of the drive and stipulated that the numerical targets were not to be exceeded without authorization of the Moscow headquarters of the NKVD.

Nothing indicates that the operation enjoyed a more orderly implementation than any other campaign in the Soviet system of planning. Available documentation on the course of the action is fragmentary, but it shows that after mid-February 1938, when according to the initial orders the operation should have been over for more than two months, the chief of the NKVD requested additional funding for the detention and transportation of about twice the number of people spoken about in the original directives. Moreover, the “Special Troikas” had largely “overfulfilled plans” by this time, having doomed 688,000 people before the end of 1937. Siinilarly, the expectations of the NKVD boss proved equally low compared to the 413,433 persons actually subjected to the jurisdiction of the local “troikas” in 1938. Local enthusiasm outstripped the expectations of the center.

In general, the leadership of the terror was not very good at predicting events. In December of 1936, NKVD chief N. I. Ezhov issued a secret order to the effect that the number of inmates at SEVVOSTLAG (Kolyma) should be 70,000 in 1937 and 1938. (This was its population as of July 1936.) But this “plan” was overfulfilled by 20,000 in the second half of 1937, and by the end of 1938 the camp housed 138,170, twice the planned level. Characteristically, as late as February 1938, the GULAG administration was at a loss to give the exact number of victims falling under its authority nationally.

Some local camp commandants found the numbers of convicts modest by the early months of 1938 and bombarded Moscow with telegrams asking for a larger “labor force,” probably because their production plans were calculated on the basis of larger contingents than the ones at their disposal. Still, hundreds of thousands of new inmates arrived after the summer of 1937 to camps unprepared to accommodate them. At the moment when the head of the secret police was applying for an increase in the NKVD budget to receive a new influx of prisoners, reports of the procurator general-who was supposed to supervise penal institutions-painted a dreary picture of the lack of elementary conditions of survival in the GULAG system as well as of starvation, epidemic disease, and a high death rate among those already there. The year 1938 saw the second highest mortality in hard regime camps before the war and probably also in prisons and labor colonies, where 36,039 deaths were recorded, compared to 8,123 in 1937 and 5,884 in 1936.

Returning to the question of plan and control over the purge, we find a letter in which the NKVD chief promised to improve the poor camp conditions, yet he reported figures for the increase in GULAG population different from the data reported by his own administration. Evidence also suggests that the NKVD and the Central Committee issued directives during the drive that were incompatible with each other. In addition, there is at least one republic on record, that of Belorussia, where vigilant local officials continued mass shootings for a time even after an order was dispatched calling for an end to the wholesale purge

Although the theoretical capacity of the prisons in Turkmenistan was put at 1,844 places, 6,796 people had been locked up in them at the beginning of 1938, and 11,538 by May; this was clearly unanticipated in Moscow. The dimensions the campaign reached in the republic explains the over-representation of Turkmen among camp inmates. Other ethnic groups also suffered-at one time, all of Ashkhabad’s 45 Greek residents were arrested as members of an “insurrectionary organization.” The NKVD chief of the republic prescribed “control figures for cases of espionage [and] sabotage” as well as specific “limits” for the number of arrests to celebrate May Day, which suggests that after a while, the operation was farmed out to regional heads of the secret police. A fife at a factory became an occasion to meet “quotas” for sabotage by arresting everybody who happened to be there and forcing them to name their “accomplices” (whose number soon exceeded one hundred persons). If nothing else worked, it was always possible to round up people having the bad luck to be at the marketplace, where a beard made one suspect of the “crime” of being a mullah and where more than 1,200 “counterrevolutionaries” were seized in a matter of five months. Mock executions and incredibly savage torture were used in Turkmenistan to wring out confessions to all sorts of “subversive acts” and “organizations.” To be sure, neither torture nor trumped-up cases was a Turkmen monopoly: the records show that both became widespread in the wake of the wholesale purge the “Special Troikas” spearheaded.

This state of affairs illustrates the problems posed by our sources on the question of “politicals.” A person arrested for his “suspicious” Polish origin or shot because of having been married to a Pole in the past was no doubt accused of being a “counterrevolutionary.” We can also only wonder how many victims shared the fate of namesakes and were sentenced to long terms or shot as alleged former members of defunct parties. How many people were like the peasant who had been condemned “merely” to ten years but whose paperwork slipped in among that of people slated for capital punishment? (He was shot with them.)

Probably, most such people figure in our data on “politicals,” even if some of the mistakenly executed were listed under the heading of their original “non-political” sentences.

Last but not least, there was the purge of the purgers: how “counterrevolutionary” were the great number of officials of the NKVD and the judiciary who were denounced for “anti-Soviet activities” after November 1938, when the Central Committee abolished the “troikas,” called off the purge, and decided that “enemies of the people and spies having made their way” into the secret police and the procuracy had been responsible for the terror of the preceding period ? Many of these “hostile elements”were sentenced as “politicals,” just as the majority of those they had cruelly mistreated, although they continued to protest their fidelity to the regime until the very end.

But whatever we think about “counterrevolutionaries,” their identified cohort constituted 34.5 percent of the camp population by 1939. This was not their largest share in the pre-war period: at the beginning of1932, people sentenced for “political” reasons in what corresponded then to hard regime camps comprised 49 percent of the inmates. The widespread recourse to capital punishment in 1937-1938 is responsible for holding the proportion of “counterrevolutionaries” under 50 percent until 1946. The percentage then declined again, probably as the result of a renewed offensive against pilferers of public property. If we superimpose the numbers of purportedly political inmates on the oscillating population of the labor camps from year to year, we find that while the proportion of “counterrevolutionaries” fluctuated, their aggregate numbers remained remarkably constant from 1939 until Stalin’s death (Figure B). This suggests that, numerically, a cohort of “politicals” was raken into the camps at the time of the Great Terror and remained relatively constant in future years.

The time of the great purges (19361939), as Figure C indicates, was numerically not the period of greatest repression. even if we take into account the masses of people shot in 1937-1938 and the much less frequent recourse to capital punishment from the late 1940s. Annual numbers of detainees were greater after World War II, reaching a peak shortly before Stalin’s death. If we extract the war years from the trend, we find that the picture is one of steadily increasing repression throughout the 1930s and 1940s.

Looking specifically at the hard regime camp populations (Figure C and the Appendixes), we find that in the twenty years from 1934 through 1953, the annual population increased in fourteen of the years and dropped in six. Of the six declining years, four were wartime; we know that approximately 975,000 GULAG inmates (and probably also a large number of persons from labor colonies) were released to military service. Nevertheless, the war years were not good ones for the GULAG. First, many of those released to the army were assigned to punitive or “storm” formations, which suffered the heaviest casualties. Second, at the beginning of the war, prominent political prisoners were transferred and isolated in the most remote and severe camps in the system and most “politicals” were specifically barred from release to the military. Third, of the 141,527 detainees who had been injails and evacuated during the first months of the war from territories soon to be occupied by the enemy, 11,260 were executed. Fourth, in the first three years of the war, 10,858 inmates of the GULAG camps were shot, ostensibly for being organizers of underground camp organizations.

Finally, wartime life became harder for the remaining camp residents. More than half of all GULAG deaths in the entire l934-1953 period occurred in 1941-1943, mostly from malnutrition. The space allotment per inmate in 1942 was only one square meter per person, and work norms were increased. Although rations were augmented in 1944 and inmates given reduced sentences for overfilhng their work quotas, the calorie Content of their daily provision was still 30 percent less than in the pre-war period. Obviously, the greatest privation, hunger, and number of deaths among GULAG inmates, as for the general Soviet population, occurred during the war.

The other years of significant population decrease in the camps were 1936 and 1953-1954. In 1936, the number of persons in both the GULAG system and labor colonies declined, as did the proportion of those incarcerated for “counterrevolution” and on sentences of the NKVD. Similarly, while the aggregate numbers of detainees were generally increasing between 1934 and 1937, the rate of increase was falling. In 1953, the year that saw the deaths of both Stalin and his secret police chief L. P. Beria, more than half of the GU LAG inmates were freed.

We have fairly detailed data about the internal movement of persons-arrivals, transfers, deaths, and escapes-inside the strict regime camp network (see the Appendixes and Figure D). They confirm Solzhenitsyn’s metaphor that this was a universe in “perpetual motion.” Large numbers of persons were constantly entering and leaving the system. During the 1934-1953 period, in any given year, 20-40 percent of the inmates were released, many times more than died in the same year. Even in the terrible year of 1937, 44.4 percent of the GULAG labor camp population on January 1 was freed during the course of the year. Until 1938-1939, there were also significant numbers of escapes from the hard regime camps. In any year before 1938, more of the GULAG inmates fled the camps than died there. A total of about 45,000 fugitives were on record in the spring of 1934, a year when a record number of 83,000 detainees took flight. Between 1934 and 1953, 378,375 persons escaped from the GULAG camps. Of them, 233,823 were recaptured, and the remaining 38 percent made good their escape. The data show, however, that the number of escapes fell sharply beginning in 1938, as Stalin with Ezhov and then with Beria tightened camp regimes and security.

The data also indicate that the average length of sentence increased in the last years before the war. The longer terms “counterrevolutionaries” were likely to

receive must have contributed to the growth of the proportion of people serving more than five years. However, Table 9 suggests that despite a notable drop in the share of long terms meted out by the courts-the sentencing policy for inmates of hard regime camps came closer by the late 1930s to the one applied to “politicals” around mid-decade.

Even if most camp convicts were “non-political,” were only serving sentences of up to five years, and hundreds of thousands were released every year, the GULAG camps were horrible places. Work was hard, rations were barely adequate, and living conditions were harsh. The inmates were exposed to the exactions of fellow prisoners and especially to the cruelty of the guards. Behind our figures lies the suffering of millions of people.

Table 9. Lenght of Sentences during Stalinist Repression, 1935-1940 (by percent)

Length:	RSFSR courts for Common crimes, First half of:		USSR civilian courts for political crimes, first quarter of:	USSR courts	In GULAG camps, January
	1935	1936	1936	1939	1940
10+ years	-	-	-	0.1	1.0
5-10 years	20.0	17.6	50.7	4.0	42.2
Up to 5 years	80.0	82.4	44.2	95.9	56.8

The long-awaited archival evidence on repression in the period of the GreatPurges shows that levels of arrests, political prisoners, executions, and general camp populations tend to confirm the orders of magnitude indicated by those labeled as “revisionists” and mocked by those proposing high estimates. Some suspicions about the nature of the terror cannot be sustained, others can now be confirmed. Thus inferences that the terror fell particularly hard on non-Russian nationalities are not borne out by the camp population data from the 1930s. The frequent assertion that most of the camp prisoners were “political” also seems not to be true. On the other hand, the new evidence can support the View, reached previously by statistical study and evidence of other types, that the terror was aimed at the Soviet elite. It also confirms the conclusions of authors who had studied the available sources and shown the uncertainties of legal theory and penal practice in the 1930s. In addition, it seems that much of the process was characterized by high-level confusion and by local actions in excess of central plans.

The Stalinist penal system can be profitably studied with the same sociological tools we use to analyze penal structures elsewhere. It contained large numbers of common criminals serving relatively short sentences, many of whom were released each year and replaced by newly convicted persons. It included a wide variety of sanctions, including non-custodial ones. For most of those drawn into it, it was in fact a penal system: a particularly harsh, cruel, and arbitrary one, to be sure, but not necessarily a one-way ticket to oblivion for the majority of inmates.

Yet it is also important to highlight three specific features. For the first, the use of, capital punishment among the “measures of social defense” sets Soviet penal practices apart from those of other systems, even though the number of executions shows a sharp decrease after the dreadful dimensions in 1937-1938. Second, the detention system in the second half of the 1930s (and perhaps at other times) was directed against educated members of the elite. Third, it had a clearly political purpose and was used by the regime to silence real and imagined opponents.

Our attempt to examine the repression of the Stalin period from the point of view of social history and penology is not meant to trivialize the suffering it inflicted or to imply that it was “no better or worse” than in other authoritarian states. Although repression and terror imply issues of politics and morality, above all for those who perpetrate or justify them, we believe that scholars can also study them as a question of historical precision. The availability of flew data permits us to establish more accurately the number and character of victims of the terror and to analyze the Stalinist repressive system on the basis of specific data rather than relying on the impressions and speculations of novelists and poets. We are finally in a position to begin a documented analysis of this dismal aspect of the Soviet past.

A Note on Sources

The GARF (TsGAOR) collection we used was that of the GULAG, the Main Camp Administration of the NKVD/MVD (the USSR Ministry of the Interior). This collection consists of nine inventories (opisi), the first of which, that of the Secretariat, contains the main body of accessible data on detainees. To be sure, it was not possible to scrutinize the more than 3,000 files of this opis', so we restricted ourselves to those that promised to tell the most about camp populations.

Accurate overall estimates of numbers of victims are difficult to make because of the fragmentary and dispersed nature of record keeping. Generally speaking, we have runs of quantitative data of severM~types: on arrests, formal charges and accusations, sentences, and camp populations. But these “events” took place under the jurisdiction of a bewildering variety of institutions, each with its own statistical compilations and reports. These agencies included the several organizations of the secret police (NKVD special tribunals, known as troikas, special collegia, or the special conference [osoboe soveshchanie]), the procuracy, the regular police, and various types of courts and tribunals.

For example, archival data on sentences for “anti-Soviet agitation” held in different archival collections may or may not have explicitly aggregated such events by the NKVD and the civilian courts. Summary data on “political” arrests or sentences may or may not explicitly tell us what specific crimes were so defined. Aggregate data on sentences sometimes include persons who were “sentenced” (to exile or banishment from certain cities) but never formally “arrested”; when we compare sentencing and arrest data, therefore, we do not always have the information necessary to sort apples from oranges. Similarly, our task is complicated, as shown above, by the fact that many agencies sentenced people to terms in the GULAG for many different types of crimes, which were variously defined and categorized. We believe, however, that despite the lack of this information, we now have enough large chunks of data to outline the parameters and to bring the areas for which we lack data within a fairly narrow range of possibility.

Further research is needed to locate the origins of inconsistencies and possible errors, especially when differences are significant. We must note, however, that the accuracy of Soviet records on much less mobile populations does not seem to give much hope that we can ever clarify all the issues. For instance, the Department of Leading Party Cadres of the Central Committee furnished different figures for the total party membership and for its ethnic composition as of January 1, 1937, in two documents that were nevertheless compiled about the same time. Yet another number was given in published party statistics. The conditions of “perpetual movement” in the camp system created even greater difficulties than those posed by keeping track of supposedly disciplined party members who had just seen two major attempts to improve the bookkeeping practices of the party.

At times, tens of thousands of inmates were listed in the category of “under way" in hard regime camp records, although the likelihood that some of them would die before leavingjail or during the long and tortuous transportation made their departure and especially their arrival uncertain. The situation is even more complicated with labor colonies, where, at any given moment, a considerable proportion of prisoners was being sent or taken to other places of detention, where a large number of convicts served short terms, and where many people had been held pending their investigation, trial, or appeal of their sentences. The sources are fragmentary and scattered on colonies, but it seems that A. N. Dugin’s attempt (see the Appendixes) to find figures for the beginning of each year - which was checked by V. N. Zemskov - yielded rather accurate results. Even so, we are not certain that errors have not slipped in.

Moreover, we do not know at the time of this writing if camp commandants did not inflate their reports on camp populations to receive higher budgetary allocations by including people slated for transfer to other places, prisoners who were only expected to arrive, and even the dead. Conversely, they may have reported low figures in order to secure easily attainable production targets.

We made extensive use of a series of statistics that were compiled about 1949 and that followed the evolution of a great number of parameters from 1934 tip to 1948. We indicated some instances in which current periodic reports of the accounting department furnished slightly different figures from those of 1949 (see the notes to Tables 3, 4, and 6) and one case in which an NKVD document in 1936 gave data similar to but not entirely identical with those calculated after the war (note to Table 8). In these as well as in most other instances, the gaps are insignificant and do not call into question the orders of magnitude suggested by the postwar documents, whose figures are, as a rule, somewhat higher than the ones recorded in the 1930s. A notable exception concerns escapes, because a 1939 report mentioned almost twice as many fugitives for 1938 as the relevant table of 1949. Although we have no explanation for this discrepancy at this moment, we can speculate that the fact that a 1939 medical report showed lower mortality figures in hard regime camps in the years between 1934 and 1939 than the 1949 account may be because the latter also includes people who had been executed.

Another source we relied on consists of four tables concerning people arrested and sentenced “on cases of the secret police” from 1921 through the first half of 1953. A peculiarity of the document is that while enumerating sentences and arrests up to 1938, it lists fewer people arrested in 1935 and 1936 than sentenced. All the while quoting the same figure for 1935 detentions as does our source, a letter signed by the head of the NKVD also speaks of more persons against whom “proceedings [had been] instituted” than those arrested. We know that some of the victims of the “cleansing” of border zones and major urban centers of “socially alien elements” had been arrested before being bankhed to faraway localities, although most of them seem to have been exiled without arrest by decisions of the NKVD jurisdiction. We also have information in this period about defendants in affairs of “anti-Soviet” agitation who had been left free pending their trial, as well as instances of the judiciary asking the police to “resolve by administrative order” cases in which there was no legal ground for conviction, a good many of which were not necessarily initiated by the NKVD.

We cannot stress enough the fact that this is only the first exploration of a huge and complex set of sources; little more than scales, ranges, and main trends of evolution can now be established. Although the above-mentioned circumstances cannot guarantee exactitude, there are good reasons for assuming that the data are reliable on the population of strict regime camps, on orders of magnitude, and on the general orientation of penal policy. There is a remarkable consistency in the way numbers, from different sources, evolve over the period under study and a notable coherence among the figures to which different types of documents refer at particular moments.

Moreover, figures produced by researchers using other archival collections of different agencies show close similarities in scale. Documents of the People’s Commissariat of Finance discuss a custodial population whose size is not different from the one we have established. In the same way, the labor force envisioned by the economic plans of the GULAG, found in the files of the Council of People’s Commissars, does not imply figures in excess of our documentation. Last but not least, the “NKVD contingent” of the 1937 and 1939 censuses is also consistent with the data we have for detainees and exiles.

Note: The 1938 data for the population of colonies also includes prison inmates, who numbered 548,417 on February 10, 1938, and the 1946 population, which contains 444,500 persons sentenced to “corrective work” without detention; GARF (TsGAOR), fond 9414, opis’ 1, delo 330, listy 55; d. 1139, l. 88; d. 1259, l. 18. The figure 1950 for “politcals” includes detainees in labor colonies. Camp and colony data are unavailable for December 31, 1953 and are here replaced by numbers for April 1, 1954, when 448,344 “counterrevolutionaries” were held at these places of detention.

Sources: GARF (TsGAOR), fond 9414, opis’ 1, delo 1155, listy 2-3 (camps and “counterrevolutionaries,” 1934-47); d. 1190, l. 36; d. 1319, ll. 2-15, d. 1356, ll.2-3 (camps and “counterrevolutionaries,” 1948-53); f. 9413, op. 1, d. 11, ll. 1-10 (prisons); A. N. Dugin and A. Ia. Malygin, “Solzhenitsyn, Rybakov: Tekhnologiia Izhi,” Voenno-istoricheskii zhurnal, no. 7, 1991, 68-70 (for colonies: calculations verified by V. N. Zemskov on the basis of GARF (TsGAOR), f. 9414, op. 1, d. 330, l. 55). See also A Note on Sources

---

About the Authors:

.J. Arch Getty is a professor of history at the University of California, Riverside. He studied with Roberta Manning and received his Ph.D. from Boston College in 1979. He is the author of Origins of the Great Purges: The Soviet Communist Party Reconsidered, 1933-1949 (1985) and co-editor of Stalinist Terror: New Perspectives (1993). His research is on the political history of the Soviet Union in the 1930s and concentrates on the history of the Soviet Communist Party. Getty is now writing (with Gabor Rittersporn) Society and Politics in the Soviet 1930s (forthcoming, Cambridge University Press), a treatment of the state-society question in the pre-war Stalin period, and is collaborating in the editing of a series of researchers' guides to Russian archives.

Gabor T. Rittersporn is a senior research fellow at the Centre National de la Recherche Scientifique in Paris. He studied at the universities of Szeged (Hungary), Leningrad, and Tokyo, defending his doctoral dissertation at the Sorbonne in 1979. His research interests involve the interaction of collective representations, social practices, and political processes in the Soviet and post-Soviet periods, with particular emphasis on the evolution of penal policy. Rittersporn is the author of Stalinist Simplifications and Soviet Complications: Social Tensions and Political Conflicts in the USSR, 1933-1953 (1991).

Viktor N. Zemskov is a senior research fellow at the Institute of Russian History of the Russian Academy of Sciences. He received his kandidat nauk degree from the History Faculty of Moscow State University in 1974, specializing in the history of the Soviet working class. He has written The Leading Force of National Struggle: The Struggle of the Soviet Working Class in the Period of Fascist Occupation of the USSR, 1941-1944 (in Russian) (1986). In 1989, Zemskov was among the first researchers admitted to the secret archives of the GULAG system, and he published a series of articles in Argumenty i fakty and Sotsiologicheskie issledovaniia on prisoners, exiles, and repatriation in the Stalin period. He is now preparing two books, one on Soviet citizens dn forced labor in Nazi Germany, 1941-1945 and another on exiles in the USSR, 1930-1960.