child mortality (under age 5) from that of changes in adult mortality (over age 5). There are reasons to suspect that the changes and explanations for these two groups are fundamentally different.
The publication in the Soviet Union of infant mortality rates for 1971-1975—showing an increase from 22.9 to 30.6 per 1,000 births, generated considerable discussion and analysis (Blum and Monnier, 1989). Publication of the infant mortality rate was discontinued by the Soviet government in the face of still worsening mortality after 1975. With glasnost, the rates were again published, with back figures given from 1980, when the rate was 27.3.
For infant (under age 1) and child (ages 1-4) mortality in most of Eastern Europe (noteworthy exceptions being Romania and Bulgaria), we have long series of data for which there is widespread consensus that registration has been adequate for many years. The data show that changes in these rates over the last decades in Eastern Europe have not paralleled those in the former Soviet Union. Throughout the period following World War II, child mortality in Eastern Europe was similar to that in the rest of Europe, except in Romania and Bulgaria. In addition, the pace of improvement has been the same (except for an increase in Romania since 1985), with no evidence of worsening infant or child mortality during the last two decades. If the above increase in the Soviet Union in fact occurred, we must seek explanations for that change that are specific to the Soviet Union and not applicable to all Formerly Socialist Economies. The divergence in pattern also emphasizes the importance of examining time trends in the former Soviet Union by republic.
In contrast with infant and child mortality, the patterns of adult mortality observed in Eastern Europe and the partial data for the Soviet Union tell a more consistent story. With regard to mortality among adult women, the levels over the last four decades have been higher in Eastern Europe than in the former Soviet Union, but the trends until 1980 were identical. Since then, trends in the two regions have diverged. Patchy data on age-specific mortality for the former Soviet Union suggest a pattern of stagnation or slow decline among most female age groups (Blum and Monnier, 1989; Eberstadt, 1993), whereas in Hungary and Poland, mortality among women aged 30-44 and 45-59 has increased slightly.
The major demographic and epidemiological puzzle of the Formerly Socialist Economies is the sustained increase in adult male mortality, which has affected those aged 30-44, 45-59, and 60-69, and remarkably began in almost exactly the same year—1964—in all countries of the region. Partial data for the former Soviet Union indicate that similar developments occurred throughout the region at the same time (Anderson and Silver, 1990; Ryan, 1982; Cooper, 1981; Eberstadt, 1993). The increases in adult male mortality continued over nearly two decades and led to a 60 percent increase among some age groups in some countries.
Cooper and colleagues (Cooper, 1981, 1983, 1985, 1987; Cooper and Sempos, 1984; Cooper and Schatzkin, 1982a, 1982b) have argued that other