Statistics of Democide
Chapter 1: Summary and Conclusions [Why Democide?...]
Other Democide Related Documents On This Site
[November 2000 note: This chapter and the related case study in Chapter 12 of Death By Government have received considerble criticism from Poles. Accordingly, I wrote an addenda in which I provide a much more detailed historical background than given here and answers the most important criticisms]
As the Red Army pounded the Wehrmacht back toward Germany's Eastern territories in October 1944, some 4,000,000 to 5,000,000 Germans fled
After the defeat of the Wehrmacht millions of Germans tried to return eastward to their former homes or areas. In the lands they passed through or returned to famine was widespread, social and health services were totally disrupted. Signs warned refugees that they would starve to death if they lingered.
For these returning Germans and those that had found havens in the Eastern territories from the war or had remained, the "greatest forcible dislocation of persons in European history"
Any calculation of this democide should include three aspects and stages in the expulsions: those Reichdeutsch or ethnic Germans killed after the Nazis were expelled from the land and before expulsion; those killed in the process of expulsion, including those dying during the transport; and those expellees dying in Germany because of their treatment during the expulsion and from the well known conditions into which they were thrown. Numerous problems face anyone trying to determine this democide, and major uncertainties of this democide are many. Just to list the major ones:
Table 7.1 also presents the estimates, consolidations, and calculations for the democide, and my attempt to compensate for these various uncertainties. The approach is one of multiple converging calculations, consolidations, and adjustments. The result may satisfy no one, but given the information available in the sources, I believe the method converges to a least unsatisfactory democide total. To do this I did three different calculations of this democide. One is for overall estimates of the toll; another is for the total of country or area specific estimates; and a third is for a regional set of estimates. The totals for these three ways of calculating democide are given on lines 81 to 82a, and consolidated to get a most probable democide (line 83) that will be discussed below.
Moreover, within each of the three sets of estimates, I calculate a possible democide figure in three ways. First, crude deficit based losses are usually calculated by taking the end of war Reichdeutsch or ethnic German population, and subtracting the number remaining in the area in 1950, and the number of expellees in Germany or other countries in the same year. This gives one measure of the overall losses, but it is crude in that population growth is ignored. Second, I subtract the number of expellees in Germany or other countries in 1950 from the number of those expelled during 1945 to 1950 to get an alternative crude measure of expulsion losses. A third measure is, of course, the consolidation of the estimates of the number of expellees that died or were killed. As possible, I then compare these three ways of determining losses and consolidate them into an overall democide. This I do for the total figures and for each nation or region. However, these democide calculations for each nation are only tentative. As mentioned, I sum the nation-specific totals to get an alternative measure of overall democide, which I then consolidate into an overall total. Once this is calculated, I proportionally readjusted the nation-specific totals to get a final nation-specific democide.
To see this in detail, turn first to the set of overall expulsion estimates (lines 3 to 77), starting with the estimates of the pre-expulsion German population (lines 4 to 7) and their consolidation (line 9). Next I show an estimate of those deported (line 12) and then estimates of the remaining German population in 1950 and their consolidation (line 18). Following this I calculate a crude deficit for the German population (line 21). The deportations are not included in this calculation, since the population estimates (line 9) are for a period after which the deportations were mainly concluded.
Shown separately is an estimate of those that perished in the German evacuation and flight before wars end. This I give for information only, since the consolidated population estimate (line 9) is for when the war ended and presumably takes this into account.
Then I show (lines 29 to 40) and consolidate (line 41) estimates of the number expelled, that fled rather than be expelled, or were "transfer-red" (a euphemism in the sources for expulsion). These are followed by estimates and their consolidation of the number of expellees in Germany (lines 44 to 57) and other countries (lines 60 to 64). I can calculate a crude measure of German losses by subtracting the number of expellees in Germany and elsewhere in 1950 from the number expelled during the years 1945 to 1950 (line 67).
Finally, there are estimates of the actual deaths (lines 70 to 77), which are also consolidated (line 78). The other two ways of calculating deaths are also given (line 79 to 80). I then consolidate these three overall estimates into an alternative (#1--line 81) overall democide figure, which I calculate as the average of the three.
But there are still two other ways of calculating the overall toll, as mentioned previously. Shown just below the first alternative (line 82) is the result of calculating the total democide from national estimates. There is no need to go into each of these, but by way of illustration those for Czechoslovakia (lines 105 to 191) should be looked at in detail. As I did for the overall estimates, I first give the German population estimates and consolidation (lines 107 to 115) and then those for the number of Germans remaining in the country in 1950 (lines 118 to 128); there were no significant deportations reported in the sources for 1950 or after. The crude deficit is then calculated (line 131). Next I show an estimate of German war/flight evacuation dead (line 134), but since it is for a period previous to the population estimates it is not used in the calculations. I give next the estimates and their consolidation for expellees (lines 139 to 157) and those surviving in Germany (lines 160 to 167) and elsewhere (line 170). From these I calculate crude German losses (line 173). Finally, I list estimates of the number dead and then consolidate (line 187) and compare them to the other two ways of calculating the toll (lines 188 to 189). These I then average (line 190).
The result is a tentative total that I then sum with the similarly derived tentative totals for Hungary, Poland and Yugoslavia (Rumania has a zero range, as will be discussed) to get the alternative overall democide total previously mentioned (line 82). However, when it along with the other two tentative figures are consolidated, the final democide figure (line 83) is changed. This means that the tentative national totals no longer sum to the total democide. To adjust these figures consistent with the overall democide (in probabilistic terms, to now recalculate a more probable democide for the larger sample of overall democide results), I proportionally readjust the tentative total to get a final democide for each nation. For Czechoslovakia, this would mean the final democide (line 191) is calculated as (line 191) = [(line 83) x (line 190)] / (line 82). That is:
(nation's democide) = [(overall democide) x (tentative total)] / (sum of country tentative totals)
My treatment of Hungary (lines 209 to 281), Rumania (lines 448 to 506), and Yugoslavia (lines 509 to 570) in the table is similar. Since the tentative toll for Rumania is zero, however, the tentative total is the final democide.
Some comment is necessary for the expulsions from new Poland (including old Poland and the former German Eastern territories and Danzig). The sources give various figures for "Danzig", the "Eastern territories", "Oder-Neisse territories", "east of the Oder-Neisse rivers" or "line", and "Poland" (which can be old Poland prior to September 1939, or new, post-war Poland). Consequently, I separated and compartmentalized these estimates depending on what is written or implied by their author. First, those estimates referring to Danzig alone are listed separately (lines 193 to 206). Then those estimates clearly referring to all or most of Germany's Eastern Territories (East Prussia, Eastern Pomerania, Eastern Brandenburg, and Silesia, and whether or not Danzig is also included) are grouped (lines 284 to 355), and for which I make a preliminary total calculation (line 359). But there is as well old Poland from which Germans were also expelled. I listed and consolidated the relevant estimates in the table and made the usual calculations (lines 362 to 398), from which I determine a preliminary total (line 399).
Next I show the estimates and consolidations for new Poland (lines 402 to 440), from which I calculate the tentative expulsion dead (line 443). Now new Poland comprises Germany's former Eastern areas and old Poland (minus the Polish eastern territory taken by the Soviets). So I can add my tentative total for each of these (lines 359 and 399) to get an alternative measure of new Poland's expulsion dead (line 444). Since there is no reason to weight one way of calculating the democide over the other, I averaged both to get the tentative total (line 445) for new Poland. This I then add to the other nation totals to get the overall alternative democide (line 82). As I did for Czechoslovakia, this overall democide now enables me to adjust proportionally the Polish total (line 446) to finally get Poland's democide. From this the tentative totals for old Poland and German's Eastern territories can also be proportionally adjusted to get their final democide figures (lines 360 and 400).
To return to the overall democide calculations, I calculate the third alternative democide total (line 82a) by using estimates of the expulsion toll for a group of Eastern European nations, which are the Baltic States including Memel, old Poland, Hungary, Rumania, and Yugoslavia (lines 574 to 576). I add the consolidation (line 577) of these estimates to the tentative totals for Czechoslovakia (line 190) and Germany's Eastern areas (line 359) to get this third alternative. The tentative totals have to be used here, since to have added the final democide figures would have been circular (they are calculated as proportions ultimately based in part on this very third alternative).
Looking now at the three alternatives (lines 81 to 82a), the one based on national estimates is much lower than the other two. To make sure in this final accounting that the low and high are indeed lower and higher bounds on the democide, for the democide range (line 83) I take the lowest and highest among the democide alternatives. The most prudent way of determining the mid-value is to calculate the average of the three alternative mid-values, which I did. This gives a mid-value very close, in fact, to the average--1,803,500--of all nine alternative figures (on lines 81 to 82a). Finally, we have a range (line 83) of 528,000 to 3,724,000 deaths as the human cost of these expulsions. Most probably 1,863,000 died.
Compare this final, overall democide figure to the death estimates from the sources (lines 70 to 77). It brackets all these estimates of the total dead and also all are above the final mid-value. This clearly indicates that the final democide figure I calculate here is conservative.
At the end of the table I calculate the overall and annual democide rates--the proportion of the Reichdeutsch/ethnic Germans expelled to potential expellees (lines 593 to 594). For the six years 1945 to 1950, over 1 out of every 10 Reichdeutsch or ethnic Germans in post-war Eastern Europe were killed or died in the process of expulsion.
Poland is of special interest because of the large number of Germans its regime killed. But these are not the only ethnics that it murdered. Table 7.2 recapitulates the total democide of Reichdeutsch and ethnic Germans (line 21) and adds to this the Polish post-war killing of Jews (lines 8 to 12) and Ukrainians (lines 15 to 17). I suspect these numbers are very conservative and do not take into account the murder of war-time collaborators; nor the execution of captured guerrillas than fighting the pro-Soviet Polish regime, and guerrilla sympathizers. But I can find no estimates of such a toll in the sources. In any case, the total democide for these years (line 24) still well places Poland among the megamurderers.
* From the pre-publisher edited manuscript of Chapter 7 in R.J. Rummel, Statistics of Democide, 1997. For full reference to Statistics of Democide, the list of its contents, figures, and tables, and the text of its preface, click book.
1. de Zayas (1979, p. xx).
2. Whiting (1982, p. 198).
3. de Zayas (1979, p. 103).
4. G. C. Paikert quoted in Buhler (1990, p. 103). Although this quote refers to the expulsion of the German population from the Oder-Neisse territories, it must apply even more forcibly to the expulsions from all of Eastern Europe.
5. Wiskemann (1956, p. 91).
6. From an unattributed quote in Zielinski (1961, p. 57).