The Effects of the Registration System on the Seasonality of Births: The Case of the Soviet Union

Scientists study seasonality in order to understand the effect of environmental, biological, and social factors on demographic events. Poor data quality can also affect seasonal variation in mortality and fertility statistics. The influence of error on seasonal data becomes crucial as researchers analyze timing and spacing of events in time-series analyses. In this study we examine the reported number of births by month in the Soviet Union for 1950, 1955, and 1958-85. The lowest number of births has typically occurred in December, and the highest in January. This seasonal pattern is not consistent with any plausible biological or behavioural explanation. It is probably an artifact of attributing births that actually occurred during the preceding December or earlier, to January. This implies that Soviet statistical practice has not followed the stated policy of attributing births to the period (day, month, year) in which they occur. A substantial reduction between the 1950s and 1985 in the December-January peak in reported births implies marked improvement in Soviet vital registration statistics. The Soviet case shows that characteristics of the registration system can impart a particular seasonal pattern to demographic data. It also shows that officially prescribed procedures are not always followed in data generation, even in centrally planned economies. Researchers should carefully examine the quality of seasonal data before concluding that the data reflect real variations in demographic behaviour.