The Global Zeitenwende
How to Avoid a New Cold War in a Multipolar Era
FOR Western analysts of Soviet affairs who have at one time or another in the past 20 years suffered from the dearth of official statistical information on the Soviet population, the long-awaited relief has finally begun to appear. On May 10 the Central Statistical Administration of the U.S.S.R. released selected preliminary results from the All-Union Census of Population of January 15, 1959, with the implication that there will be more to come (although how much more is not clear), following the completion of tabulation toward the end of the year.
From the data so far released one fact stands out. Soviet loss of life in World War II was far greater than we had supposed. Indeed the deaths among the male population simply stagger the imagination. Before developing this point, let us glance at the data so far released and discuss briefly their validity.
The preliminary census results show a total population of 208,800,000, together with the following breakdown:
In addition, we are given the urban and rural population for each of the 15 republics and their constituent administrative subdivisions (oblasts, krais and okrugs); the number of urban areas of different types, and their population; and the population of cities which are centers of administrative divisions and subdivisions as well as of cities with more than 50,000 persons. Also enumerated by the census, but not reported in the preliminary results, are the categories of age, nationality, language, education, family status, social classification, source of livelihood and occupation by branch of industry.
The 1959 census of population is the seventh planned since the beginning of the Soviet period, the sixth to be carried out in one form or other, and the fifth from which at least selected data have been published. The first All-Union Census of 1920 was conducted under abnormal conditions and the results were consequently incomplete and of generally poor quality. The urban census of 1923 and the All-Union Census of 1926 were published in full. A census was planned for 1932, but never carried out, and a census was carried out in 1937, but the results were never published. The results of the 1939 census (the most recent prior to that of 1959) were published on a very restricted scale, allegedly because of "wartime conditions."
The statistical validity of the 1959 census is difficult to establish on the basis of the preliminary results, and a careful examination of this question must therefore be postponed. However, judging by the scale of organization and preparation and the various devices introduced to enhance the internal consistency and reliability of the results, the latest census will probably hold up well in comparison with censuses of other countries.
The additional question, often raised in Western circles, of whether the Soviets deliberately "manipulate" the published figures is extremely difficult if not impossible to answer. Since outward signs point to a certain degree of reliability of the results generated by the census itself, the possibility of "manipulation" for publication would have to be tested by a painstaking search for inconsistency with other data, and for mutual inconsistency within the published results themselves. This has not been attempted here. It has, however, been the experience of most specialists that the Soviets do not deliberately alter data in cases where categories are clearly identifiable, although the categories themselves are in many instances quite obscure; and that the withholding of data or their publication in obscure form seems to be the preferred method of operation.
A total Soviet population of 208,800,000 as late as 1959 reveals, as we have said, an absolutely incredible loss of life during World War II. It is true that we were alerted to this possibility through the publication of an official estimate of 200,200,000 as of April 1, 1956. This total, only slightly higher than that for 1940-41, was based on non-census sources of information which were not without admitted statistical weaknesses. A number of Soviet statisticians felt that the 1959 census would show, if anything, a somewhat higher total population, and therefore a less awesome loss of life during the war. The fact is that the total population reported from the 1959 census is approximately 1,400,000 less than that implied by the 1956 estimate, taking into account the natural increase of the population reported for the intervening months.
The "war losses" which the census figures suggest are difficult to measure with accuracy: first, because an effort to recreate what would have happened if there had been no war is subject to widely varying hypotheses with respect to the birth rate; and second, because the total effect of the war on the growth of the population takes several generations to have its full effect. Nevertheless, if we confine ourselves to the period of the war and the years immediately following, and make simplifying assumptions with respect to birth and death rates in the absence of war, the sum total of persons who would not have died plus those who would have been born is in the neighborhood of 45,000,000.
This staggering total of population losses stems from the following considerations:
(1) As of January 1, 1940, the population on territory comparable to that of postwar Russia but excluding persons who subsequently emigrated from the U.S.S.R., may be estimated as 194,000,000.[i]
(2) Ten years later, as of January 1, 1950, the population may be estimated as 179,000,000. Thus, the population is estimated to have declined (other than through emigration and territorial adjustment) by 15,000,000 between 1940 and 1950. (The population decline during the war years themselves was more drastic, from almost 200,000,000 on July 1, 1941, to some 170,000,000 in 1945.)
(3) If there had been no war, the population of 194,000,000 in 1940 would have reached a total of about 224,000,000 in 1950. This assumes a relatively sharp decline in birth rates from the high prewar level of 38 per thousand reported for 1938--an assumption which is consistent with the actual levels reached by 1950 and also with Lorimer's hypothetical calculations based on the experience of other developing countries. Mortality rates in the absence of war are also assumed to have declined between 1940 and 1950, following the pattern of other countries.[ii]
Compared to the actual total population of 179,000,000 in 1950, therefore, a hypothetical total of 224,000,000 implies "war losses" of 45,000,000. These may be distributed between the number who would have been born and the number who would not have died, as follows:
(1) The actual population aged 0-9 in 1950 may be estimated as 30,000,000, whereas, if there had been no war, the population in this group would have been approximately 50,000,000. This places the number of persons who were not born or who died in infancy as a result of the war at 20,000,000, and represents a decline in the birth rate and/or an increase of the infant mortality rate of more than 50 percent during the war years themselves.
(2) By subtracting the 0-9 age group from the population total for 1950, we obtain a figure of 149,000,000 aged 10 years and over, which represents the postwar survivors of the 194,000,000 persons alive on January 1, 1940. That is to say that some 45,000,000 persons actually died in the decade prior to 1950, compared to about 20,000,000 who would have been expected to die on the basis of the hypothetical mortality rates set forth above. In other words, the war was responsible for the death of some 25,000,000 persons out of the population alive in 1940, to which must be added the 20,000,000 who were not born or did not survive infancy as a result of the war.
War losses totalling 45,000,000 seem all the more staggering when they are distributed by sex. The 1959 census reports 114,800,000 females and 94,000,000 males, indicating an absolute deficit of 20,800,000 males. This may be compared to a deficit of 700,000 according to the 1897 census, 5,000,000 in 1926 and 7,200,000 in 1939. The increases in the male deficit compared to the pre-Soviet period reflect, successively, World War I and the Civil War, the period of collectivization and the food shortages of the early 1930s, and World War II.
At the same time, the amount by which the male deficit increased due to World War II is considerably beyond that which might have been expected. From several partial references and surmises, a relatively liberal allowance for the number of males killed in military service as a direct or indirect result of the war is approximately 10,000,000. This allows for the death of approximately one-half of the 20,000,000 men reported to have served at one time or another in the armed forces of the Soviet Union during World War II.
As shown above, an estimated 25,000,000 persons died from all causes between 1940 and 1950, over and above the number who would have died if there had been no war. If 10,000,000 died in uniform, we must account for civilian losses of 15,000,000 (exclusive of infant mortality). From the data on the number of males and females from the 1959 census, it follows that of these 15,000,000 excess civilian deaths, some 11,000,000, or three-quarters, were males, and 4,000,000 females.
What cause or combination of causes can explain why three times as many males as females died outside the military service during and after the war, over and above the number who would have died under "normal" conditions? Since it is most unlikely that military deaths were greater than the estimated 10,000,000, the explanation would entail the following possibilities: (1) the population of areas annexed by the Soviet Union after the war included a smaller ratio of males than generally assumed; (2) the emigrant population included a relatively high proportion of males; and probably most important, (3) the various factors accounting for excess civilian deaths--military action against civilians; the general lowering of public health and nutritional standards; the use of labor battalions and population extermination by the Germans; and the use of correctional labor camps by the Soviets themselves--fell most heavily on males.
Perhaps there is a useful analogy in the experience of the 1930s. During this period the male deficit increased by more than 2,000,000--from a total of 5,000,000 in 1926 to 7,200,000 in 1939--whereas under "normal" conditions the deficit should have declined. In other words, the "excess civilian deaths" of the 1930s, at a time of mass collectivization, large-scale population migration from rural to urban areas, and the widened use of correctional labor camps, also fell more heavily on the male population.
For nearly a decade, the male deficit has been declining and, according to the census results, is now confined to the population aged 32 and older. Therefore, the problem created by the shortage of males in the labor force is already lessening. On the other hand, the effect of the war on the age group now entering the labor force is only beginning to be felt. Whereas the population of working ages has normally increased by 1,000,000 to 2,000,000 each year, during the next decade the growth will be only a few hundred thousand annually. Growing attention to the implications of this development is shown by recent changes in Soviet policies in regard to the recruitment, training and utilization of manpower.
The pattern of Soviet population growth which emerges from the preliminary census results, together with other data, must be examined from several points of view. If we consider the period as a whole since the start of the industrialization drive--that is, from the 1926 census to the 1959 census--the average rate of population increase is approximately 0.8 percent per year.[iii] This is lower than the long-run average which includes the Imperial Russian period, because from 1850 to the present, on Soviet territory, the population increased on the average by slightly more than 1.0 percent per year, and in the fifty-year period prior to 1850, on Imperial Russian territory, by slightly less than 1.0 percent per year.
One reason for the lower rate of increase over the past three decades, of course, lies in the demographic effects of World War II, described above. This is illustrated by the fact that between the 1939 census and 1959 census the average rate of population increase was only 0.5 percent per year. At the same time, the peacetime rate of population increase since the start of the industrialization drive has also been less than that for earlier years. In the first years of rapid industrialization (1931-1934), the decline in the rate of population growth was nothing short of catastrophic, for reasons to which we have already alluded. Although population growth recovered by the later 1930s, the impact of these few early years was severe enough to reduce the average increase for the period 1926-39 to 1.23 percent per year, or only a little more than one-half of the rate that Soviet planners had anticipated. Nevertheless, this is a higher rate than many other countries attained in these years.
The rate of population growth since 1950 has been about 1.6 percent--also less than that before the start of the industrialization drive, but the difference here is not very great; and, again, as Soviet sources emphasize, a rate of 1.6 percent per year is relatively high by comparison with most other countries. However, this relatively moderate decline in the rate of population growth under peacetime conditions conceals a relatively sharp decline in both birth and death rates since the early 1920s. Certain vital statistics published with the preliminary census results, together with data from other sources, show a crude birth rate of about 25 per thousand now, compared to more than 40 per thousand on the eve of the industrialization drive; a death rate of less than 8 compared to 18; and average life-expectancy of 67 years compared to 44 years. Standardization for changes in age structure compared to the 1920s would have the effect of lowering the present birth rate and raising the death rate somewhat, so that each is now about one-half that of 35 years ago. Approximately one-half of the decline in the crude death rate from 18 to 8 per thousand is due to the decline in infant mortality, which is reported for 1957 at 45 deaths per one thousand births, compared to 184 in 1940.
A crude death rate of less than 8 per thousand is below that of almost any other country and compares with a rate of less than 10 in the United States, which has a higher proportion of old people. An infant mortality rate of 45 per thousand births equals the rate in this country about 1940.
Only one recent Soviet source appears to include any attempt to explain the decline in the crude birth rate.[iv] Even in this case, the prewar dates are selected for comparison so as to minimize the extent of the decline, which is then ascribed to (1) the lower rate of infant mortality (leading, we can presume, to the need or desire to have fewer children), (2) the increase in the proportion of urban residents, and (3) changes in the sex ratio resulting from wartime population losses. In all probability these factors fall considerably short of explaining the full extent of the decline. In any event, it seems quite evident that urban and probably also rural families are getting smaller--a conclusion that corresponds to the impressions of recent visitors to the Soviet Union. Forced to speculate on the underlying reasons for this turn of events, we can at least suggest a number of factors: the rapid rate at which the educational level of the population has been raised; the larger share of work falling on women, in the cities as well as in the countryside; and the continuation of the shortage of urban housing.
As the "war babies" pass through the child-bearing age in the next decade or so, the birth rate may fall to 20 per thousand or less, even if fertility rates do not decline. In other words, the birth rate will remain constant or rise only if there is a substantial increase in fertility rates.
The data on the rural and urban population reproduced above show a larger proportion of the population in urban areas (48 percent) than would have been expected by projecting the official non-census estimate as of April 1, 1956 (43 percent urban) on the basis of reported increases in the urban population between 1950 and 1955. It would seem that either (a) the urban population has in fact increased since 1956 by almost 5,000,000 persons per year, compared to 3,400,000 between 1950 and 1955, or (b) the greater accuracy of the census total tends to render the two figures essentially incomparable. In any event, the fact that almost one-half of the Soviet population now lives in urban areas places the Soviet Union among the more urbanized countries of the world, although well below such countries as the United States (59 percent), Denmark (61 percent) and the United Kingdom (80 percent).
The Soviet Union now has the same percentage of its population in urban areas as the United States had about 1910, while it began the period of rapid industrialization with about the same degree of urbanization as the United States had in 1860. However, urbanization was more rapid in the Soviet Union during the 1930s than it was in the United States after 1860. In absolute terms the urban population doubled in about the first ten years of industrialization in the Soviet Union, but in the first 20 years in the United States. Or, expressed in somewhat different terms, it took almost 20 years for the percentage of the population in urban areas to double in the Soviet Union, but almost 40 years in the United States.
These figures are explained not only by a more rapid rate of migration of the Soviet population from rural to urban areas, but by the fact that attrition in World War II fell more heavily on former rural residents in the heavily "peasant" army. This is more than enough to compensate for the fact that in the United States during the nineteenth century a relatively large proportion of the immigrant population settled in urban areas.
The preliminary census results by administrative subdivisions serve to document more accurately the postwar population shifts toward the central and eastern parts of the country. European Russia remains the most populous part of the country, but its population in absolute terms has changed very little since prewar years. Virtually the entire increase in the total population since the 1939 census, small though it is, has taken place in the areas outside of European Russia. This is accounted for by the eastward shift of industry and by the "virgin lands" campaign, both of which have entailed considerable migration. But it also reflects the fact that birth rates have remained relatively high in some of the eastward areas, particularly in Soviet Central Asia.
In summary, the significance of the census results so far released are two-fold. In the first place, they are important for what they reveal about a country passing through a major war and a rapid succession of economic and social developments; and because they document demographic factors which to some extent must be influencing the Soviet Government in formulating policy in labor, military and other fields. Secondly, these statistics derive importance from the simple fact that they have been released, to form a component of the flow of data from official sources which began in 1956 and shows promise of widening in the future.
[i] The purpose of excluding those who emigrated is to narrow the comparison with postwar data to the effect of changing birth and death rates. Sources and methods for these and other data in this article, other than from the 1959 census, may be found in W. W. Eason, Soviet Manpower: The Population and Labor Force of the US.S.R., a Ph.D. dissertation on deposit with the library of Columbia University.
[ii] The rates referred to in these assumptions are from F. Lorimer, "The Population of the Soviet Union: History and Prospects" (Geneva: League of Nations, 1946), p. 254-255.
[iii] Territories acquired since 1939 are consistently excluded from calculations made in this article.
[iv] M. Ia. Sonin, "Ob aktual'nykh voprosakh vosproizvodstva trudovykh resursov SSSR," Akademiia Nauk SSSR, Voprosy sotsialistichheskogo vosproizvodstva (Moscow: 1958, p. 257-281).