1. Barbara Anderson and Brian, Silver, “Demographic Analysis and Population Catastrophesin the USSR,” Slavic Review 44 (Fall 1985): 517, 519Google Scholar. Anderson and Silver acknowledge that “thedemographic history of the Soviet Union is marked by catastrophes…. [attributable in part to] thecollectivization of agriculture and the famine.” Like Wheatcroft, however, they avoid attributing asignificant number of these catastrophic deaths to Gulag and the Terror and favor technical adjustmentsof the sort suggested by Lorimer that reduce famine and collectivization deaths to diminutiveproportions (ibid., p. 517). Their comments about Lorimer's estimates as measures of “statisticalconsistency” rather than excess deaths should be read in this light (ibid., p. 520).

2. Lorimer estimated that there were 5.5 million excess deaths between 1 January 1929 and1 January 1939 but refrained from firmly attributing them to collectivization, Gulag, and the Terror.See Lorimer, Frank, The Population of the Soviet Union: History and Prospects (Geneva: League of Nations, 1946, p. 134 Google Scholar. Wheatcroft frequently adopts the position that most estimated excess deathshave innocent causes but has recently been willing to concede the possibility of significant, policyrelatedlosses. See Stephen Wheatcroft, “Population Dynamic and Factors Affecting it, in the SovietUnion in the 1920s and 1930s,” part 2, CREES Discussion Papers, Birmingham University (U.K.), 1976, p. 70; and idem, “New Demographic Evidence on Excess Collectivization Deaths: Yet AnotherKliukva from Steven Rosefielde?” Slavic Review 44 (Fall 1985): 508.

3. Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 519. For a discussion of the normal level of Gulag forced labor see CIA, Crime and Punishment in the Soviet Union, GC79–10010, March 1979 (Confidential); U.S. House Subcommittee on Human Rights and International Organizations of the Committee of Foreign Affairs, Forced Labor in the Soviet Union, House of Representatives and Commission on Security and Cooperation in Europe, 9 December1983, p. 103; Ger P. van der Berg, “The Stalinist System of Justice and Terror in Figures and SomeLessons for the Present Day,” paper presented at the Third World Congress for Soviet and EastEuropean Studies, Washington, 30 October to 4 November 1985; and Steven Rosefielde, “Knowledgeand Deception: Stalinist Industrialization Reconsidered,” Soviet Studies (forthcoming, April 1987).Compare Rittersporn, Gabor Tamas, “Soviet Officialdom and Political Evolution: Judiciary Apparatusand Penal Policy in the 1930s,” Theory and Society 13 (1984): 211–237 CrossRefGoogle Scholar. Anderson and Silver incorrectly assert that “ Rosefielde's main interest has been in estimating the size of the Soviet prison camp population” (Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 517). My main interest is in scientifically reconstructing the Soviet development experience and interpreting it from the standpoint of economic systems theory. They then quote Conquest to suggestthat “we are unlikely to be able to deduce with any certainty the labour camp population from anyof the demographic material presently available” (ibid., p. 519). This formulation misstates my intent.I can and do use the demographic evidence to demonstrate that my Gulag forced labor estimatesare not disconfirmed by plausible estimates of contemporaneous excess deaths. Compare Robert, Conquest, “Forced Labour Statistics: Some Comments,” Soviet Studies 34 (July 1982): 438.Google Scholar

4. Lorimer, Population of the Soviet Union, p. 119. “These assumptions are obviously arbitrary….” Compare Steven, Rosefielde, “New Demographic Evidence on Collectivization Deaths: A Rejoinder to Steven Wheatcroft,” Slavic Review 44 (Fall 1985): 51.Google Scholar. The most Anderson and Silver venture in this regard is the hollow assertion that the logic of my critique of Lorimer's adjustmentis “invisible.” Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 523.

5. Steven, Rosefielde, “Excess Mortality in the Soviet Union: A Reconsideration of the Demographic Consequences of Forced Industrialization 1929–1949,” Soviet Studies 35 (July 1983): 385–409 Google Scholar; idem, “New Demographic Evidence on Collectivization Deaths 1929–33: The Demographic Evidence,” Slavic Review 42 (Spring 1984): 83–88; idem, “New Demographic Evidence on CollectivizationDeaths: A Rejoinder to Steven Wheatcroft,” 509–516; idem, “Excess Deaths and Industrialization: A Realist Theory of Stalinist Economic Development in the Thirties,” Journal of Contemporary History (forthcoming, July 1987); idem, “Knowledge and Deception “; idem, “AnAssessment of the Sources and Uses of Gulag Forced Labour 1929–56,” Soviet Studies 33 (January1981): 51–87.

Anderson and Silver contend that Lorimer's methodology was only used as a test of statisticalconsistency, not excess deaths. This contention is both logically and factually specious. See Lorimer, Population of the Soviet Union, p. 133, “Actually, there is a discrepancy of about 5.5 million. Onepossible interpretation of this discrepancy is that it represents the magnitude of‘excess mortality’ beyond that normally expected during the period of the collectivization of agriculture.“

6. Ansley Coale's adjustment is an important and plausible component of Lorimer's more ambitiousrecomputation; see Lorimer, Population of the Soviet Union, p. 117.

7. Compare Rosefielde, “Knowledge and Deception,” table Al, with Rosefielde, “Excess Mortalityin the Soviet Union,” table 3, p. 388.

8. Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 517.

9. Ibid., pp. 523–524. The same argument is obliquely levied against my wartime natality estimates.See note 24, below.

10. Rosefielde, “Excess Mortality in the Soviet Union,” table 3, p. 388.

11. Ibid., table 2, p. 387. For a fuller discussion of the differences between these data seeRosefielde, “Knowledge and Deception. “

12. Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 524.

13. Sifman, Rosa, Dinamika rozhdaemosti v SSSR (Moscow: Statistika, 1974 Google Scholar, figure 2, p. 43. Sifman uses the cumulative fertility behavior of twenty-nine-year-old women as her proxy for thereal natality rate. Cf. Anderson and Silver, “Demographic Analysis and Population Catastrophes,” pp. 530–531. For a more detailed discussion of this issue see Rosefielde, “Knowledge and Deception. “

14. Stalin, “Otchetnyi doklad,” 1934, p. 25, cited in van den Berg, Ger P., The Soviet System of Justice: Figures and Policy (Boston: Martinus Nijhoff, 1985, p. 176 CrossRefGoogle Scholar.

15. Ibid.

16. Rosefielde, “Excess Collectivization Deaths 1929–33,” and “New Demographic Evidence,” pp. 510–511. Compare Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 523.

17. Anderson and Silver, “Demographic Analysis and Population Catastrophes,” pp. 525–526.

18. Antonov-Ovseenko, Anton, Portret tirana (New York: Khronika, 1980, p. 211 Google Scholar.

19. Antonov-Ovseenko, personal testimony.

20. Stephen, Wheatcroft, “A Note on Steven Rosefielde's Calculations of Excess Mortality inthe USSR, 1929–49,” Soviet Studies 36 (April 1984): 278 and 281Google Scholar. Also see idem, “Population Dynamic and the Factors Affecting It, in the Soviet Union in the 1920s and 1930s,” Part I, SIPSNo. 1, University of Birmingham (U.K.), 1976, p. 36. Compare Anderson and Silver, “DemographicAnalysis and Population Catastrophes,” p. 526. “Although the 1939 census probably contains errorsthe existence and magnitude of other forms of misenumeration in that census have not been demonstrated.“

21. Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 523. They simply accept Lorimer's “arbitrary” adjustment. It should also be observed that they restrict their discussion to the 1926–1927 census mortality rate, ignoring the fact that mortality rates for subsequentyears belie their preferred adjustment.

22. Ibid., table 1, column 3, p. 528; table 3, column 3, row 6, p. 531.

23. Ibid., p. 523.

24. Anderson and Silver contend that my excess death estimates for the 1940s, computed onthe assumption that the 1939 census does not contain “dead souls” are wrong because the natalityrates used for the years 1941–1945 are too high (this is the reason they assert my estimate is apopulation deficit, not an excess death statistic) and because Stephen Wheatcroft has uncoveredother shortcomings (ibid., p. 522). The first objection is refuted by Rosa Sifman's anamnestic research.See Sifman, Dinamika rozhdaemosti v SSSR, p. 43. The second is disproven in Rosefieide, “Knowledge and Deception.” Although, Soviet authorities have found it expedient to bury tens ofmillions of nonwar related excess dead in the “fog of war,” we are not required to certify theirdeception, a point properly stressed by Nekrich. See Nekrich, Aleksandr, Otreshis’ ot strakha (London: Overseas Publication Interchange, 1979, p. 158 Google Scholar.

25. This method is adopted because “a source-by-source or statistic-by-statistic approach is notthe only way to evaluate the quality of demographic estimates” (Anderson and Silver, “DemographicAnalysis and Population Catastrophes,” p. 519). This statement is true, but evaluation and disproofare two separate things. Their sensitivity tests illuminate underlying uncertainties but do not establishtheir contention that their preferred estimates are right, or that mine are wrong.

26. In this regard it should be noted that while the purpose of Anderson's and Silver's sensitivitytests is to show that my findings are out of bounds, they actually confirm them. They write: “But itis clear that one would have to make an extraordinarily optimistic assumption about the‘normal’ mortality rate, an assumption that the mortality rate was far lower than that reported for 1938–1939, to conclude that the number of excess deaths during 1927–1938 of persons who were alive in 1926was in the neighborhood of 8 or 10 million” (ibid., p. 528). If the 5.2 million adult excess deaths Icompute with the 1939 census are properly used as a standard of reference, they fall neatly withinAnderson's and Silver's range of admissibility, 3.2 million to 5.5 million adult excess deaths (ibid., p. 529). See Rosefieide, “Excess Mortality in the Soviet Union,” table 11, column 4, rows 3 and 4, p. 400. Lorimer's estimate similarly defined is 4.8 million excess fatalities. Anderson and Silver's reference to “8 or 10 million” excess deaths wrongly refers either to my estimates based on the 1939census including children or to my estimates based on the 1937 census.

27. Three additional miscellaneous criticisms warrant brief comment. Anderson and Silver assert, “Rosefieide has been inconsistent in his assumptions about the 1939 census. [He contends] thatthe 1939 census was an overcount. But earlier, when interpreting the 1939 census count of peopleworking for pay Rosefieide treated the 1939 census as if it were a‘good’ census” (Anderson andSilver, “Demographic Analysis and Population Catastrophes,” p. 526). The purported contradiction is entirely of their own contrivance. It does not necessarily follow that the census employment statistics are biased because the population aggregate is overstated. Moreover, I explicitly observedthat the census employment statistics were incomplete because they omit unpaid penal workers; seeRosefielde, “Gulag Forced Labor,” p. 73. On pp. 520–521 of their article Anderson and Silver splithairs about how best to define the concept of excess deaths. The purpose of this exercise is to imply that demographic analysis cannot support reliable assessments of fatalities caused by collectivization, Gulag, and the Terror. “We do not share Conquest's certitude about being able to estimate excess mortality from available demographic statistics” ( “Demographic Analysis and Population Catastrophes,” p. 519). Their argument is sophistic. Coroners’ reports for the victims of collectivization, famine, Gulag, and the Terror would be ideal, but estimates based on Lorimer's method and other corroborating evidence will suffice. Finally, Anderson and Silver attack the credibility of my estimates by claiming that I do not clearly grasp the distinction between population deficits and excess mortality: “But Rosefielde has interpreted population deficits as if they were excess (actual) deaths. For example, at one point he writes that‘these data indicate that the Soviet Union experienced 37 million excess deaths during the 1940s.’ In fact, the figure to which he refers is an estimate of a population deficit, not of excess deaths” (ibid., p. 522). Both the assertions, that I have not clearly elaborated the distinction between these concepts and that the figure in question is a population deficit, arenonsense. With regard to the former see Rosefielde, “Excess Mortality in the Soviet Union,” p. 385.Anderson and Silver do not spell out the rationale for their second allegation, but their text implies that they believe my natality statistics are unreliable: “It is extremely misleading to interpret population deficits as excess deaths because the population deficit includes … births that did not occur” (Anderson and Silver, “Demographic Analysis and Population Catastrophes,” p. 522). This characterizationis deceptive. It is predicated on the erroneous assertion that excess mortality estimatesare only valid if birth statistics are completely accurate. Population deficits are distinguished from excess mortality estimates not by the precision of the birth data, but by the fact that the former areprojections, whereas the latter are based on ex post facto statistics adjusted where necessary formissing observations. A glance at my computations reveals that I use ex post facto natality statistics, rather than projections, which yield excess mortality estimates as claimed, not population deficits.See Rosefielde, “Excess Mortality in the Soviet Union,” table 5, p. 393. Anderson and Silver demurbecause they believe my wartime natality rates are unreliable. Perhaps they are, but as a comparisonbetween my natality rates and those employed by Lorimer for an earlier period demonstrate thisdoes not mean that my estimates must be interpreted as population deficits; see my “Excess Mortalityin the Soviet Union,” table 5, p. 393, and table 2, p. 387.

28. See note 26.

29. Rosefielde, “Excess Mortality in the Soviet Union,” table 12. Cf. Jean-Claude Chesnais, “Some Peculiarities of Eastern Europe and Soviet Population Trends,” paper presented at the ThirdWorld Congress for Soviet and East European Studies, Washington, 30 October to 4 November 1985, pp. 6–8.