Social Mobility and Interpersonal Relations

tion, certainly for ages 50 plus 18 and possibly for ages 12-49, whether due to errors of age reporting in the 1926 19 and/or 1939 censuses or to Lorimer's expected mortality rates 1926-1939. An investigation of this problem might reveal that Lorimer's estimate of losses from collectivization should be revised. 18 Persons aged 50 and over in 1939 were aged 42-43 and over in 1931-1932. It is a reasonable conjecture that the excess mortality of males was a consequence of the terror which was primarily directed against heads of households, in connection with collectivization. number of kulaks (wellto-do peasants) who had to leave their homes ran into millions. See Eugene M. Kulischer, Europe on the Move, New York: Columbia University Press, 1948, p. 93. 19 Lorimer has stated in this connection: The corrections which Mr. Eason obtains by using sex ratios by broad age classes derived from information on literacy rates decrease the number of males (and increase the number of females) at ages 50 years and over. This improves the reliability of the distribution, and has a logical rationale. Unfortunately, in any case, the estimated numbers by five-year age classes in 1939, and in all subsequent projections, are somewhat distorted by errors in age reporting in the original 1926 census data. See Comments of Frank Lorimer, in Soviet Economic Growth, edited by Abram Bergson, 1953, page 123.