Trends and differentials in infant mortality in the soviet union, 1970-90: How much is due to misreporting?*

We use recently released data on perinatal mortality and cause of death to assess how much of the spatial and temporal variation in infant mortality in the former Soviet Union is attributable to differences in the extent of misreporting. We demonstrate that the dramatic rise in infant mortality that occurred in the mid-1970s was accounted for in large part by an increase in death rates from causes which predominate after the first month of life, particularly in the Central Asian republics, but also in the more developed Baltic and European republics. Improvements in the classification of perinatal deaths do not appear to have played a significant role in explaining trends in reported infant mortality in the 1970s, but may have been responsible for some of the rise (or lack of decline) during the late 1980s. Despite the apparent improvements in the recording of deaths that occurred shortly after birth, there is evidence in several republics of substantial misclassification of early infant deaths as late fetal deaths as recently as 1990. Because such a pattern would lead to the omission of many infant deaths, it appears that infant mortality rates may have been understated in several of the less developed republics even at the end of the period studied.