Crude death rates in the U.S.S.R.: Comment on Grandstaff's remarks

Abstract My appraisal of death rates in the U.S.S.R. goes beyond the question whether a simple correlation really measures what the critic purports to measure. Normally, the crude death rates depend upon the age structure and the force of mortality. If one persists in assuming that a reasonably high correlation between the official death rates and their independent estimates should be the main criterion of validity, the best technique will be to relate the proportion of persons aged 60 and over to the official death rates recorded for some 109 observations. The force of mortality would be approximately reflected by a given proportion of survivors to the first or second year of age. But in the 1959 Soviet Census the latter data had been concealed within the 0 to 9 age group, possibly for no other reason than to prevent the curious few from tinkering with infant mortality.