ABSTRACT
PIP: The relationship between fertility and infant mortality is interdependent. Reduction in infant mortality affects the reduction of fertility in 2 ways. The biological mechanism operates largely through lactation and the length of postpartum amenorrhea. Infant death abbreviates lactation and hastens the resumption of ovulation. The behavioral mechanism operates largely through experience with and fear of infant mortality, which motivates parents to replace lost children or to have many children as insurance against expected death. Both of these effects are dependent on prevailing social norms about ideal family size. Another behavioral explanation is that parents' sexual activity is different when an infant survives than when an infant dies. Fertility also affects infant mortality. Infant mortality has been observed to increase with the order of birth. The biological mechanisms include infant mortality increasing with the age of the mother and the health status of the mother deteriorating with successive births. Socioeconomic status also plays a role in infant mortality. An empirical study using data from 13 Indian states examines the relationship between 4 variables: percentage of higher order births (the fertility measure), infant mortality rate (IMR), per capita state domestic product, and a measure of the distribution of income. A path analysis of 2 models, one using percentage of higher order births as the dependent variable, the other using IMR as the dependent variable are carried out. Comparing the values of the path coefficients, the direct effect of infant mortality on fertility is found to be higher than the direct effect of fertility on infant mortality. The indirect effect of the development variables lead to the same observation. However, the values of the multiple correlation coefficients are small, since fertility is determined by various other factors. The relationship between fertility and infant mortality is further explored by employing a time series design using data covering 1936-1971 from Chile and Italy. The pattern revealed in the Chilean data is of declining IMR, but at high levels and a stable percentage of higher order births, fluctuating around 36-37 until 1951, after which there is a 10 year upward trend, followed by a continuous reduction after 1963. A possible explanation is that with a declining IMR, there is a simultaneous reduction of general mortality, implying better health conditions. The immediate effect is increasing the length of the reproductive span, which increases higher order births. At the initial stages of IMR decline, the association between infant mortality and fertility is weak and negative. At a later point, after 1963, both factors are declining and the correlation between them is .98. The Italian data is consistent with the later stages. Exploring the existence of a time lag in people's perceptions of declining infant mortality, the correlations indicate that the association with fertility is more with IMR of the same year rather than earlier years. The association begins operating when IMR has declined to around 110. These correlations were not subjected to tests of significance.^ieng