Abstract

PIP: Using World Fertility Survey data for Colombia, Kenya, Sri Lanka, and Pakistan, researchers tested their hypothesis that a couple's direct response to the death of a child by replacing the child differs among population subgroups. Even though the mean s for additional children differed among the 4 countries, women who had lost a child reported more of a desire for additional children than those who had not lost a child. In Colombia and Sri Lanka, the replacement effect became less strong as parity increased. On the other hand, in Kenya and Pakistan, the mean adjusted additional children wanted by those who had at least 1 child die remained high regardless of parity. The adjusted means for subsequent births showed that, in all countries, women who had lost a child were more likely to have more subsequent births than those who had not lost a child. The effect of an infant death was greatest at the early parities for women in Colombia and Sri Lanka. The probability of another birth rose somewhat in 3 of the countries, while in Kenya women at early parities were less likely to have another child. Of the 4 countries, Colombia had the shortest median length of breast feeding, so overall birth intervals and those by survival outcome in Colombia were less than those for the rening countries. In Sri Lanka and Pakistan, women who had the most education were more likely to continue childbearing than those who had no education (2 times in Sri Lanka and 3.47 times in Pakistan). Further, the probability of a subsequent birth among rural women in these 2 countries was higher than among urban women. Since many couples use contraceptives in Colombia, motivation and/or ability to have a subsequent birth was similar across all categories, e.g., education and locale.