Son Preference and Fertility in Bangladesh

Researchers followed 22819 women from Matlab in rural Bangladesh with known sex composition of living children as of July 1982 for 3.5 years. They hypothesized that the effect of son preference on fertility is stronger in a population with adequate and available contraceptive service and relatively high contraceptive prevalence than a population with little to no contraceptive use or access. From July 1982-end of 1984 the contraceptive prevalence rate in the study area climbed from 35%-46%. It ranged from 17-23% in 1983-1985 (less than national average) in the comparison area. 58% of the women had additional children in the study area compared to 69% for the control area. The larger the family size the lower the percentage of women giving birth in both areas. This relationship was especially strong in families with = or + 3 children in the study area. Both areas had an association between sex composition of children and subsequent fertility. Further in both areas the percentage having additional children at each parity level was lower when the number of living sons was higher. this preference was strongest in the study area. For example the percentage difference between women having additional children at parity 2 and no son and those with 2 sons stood at 12% (79% vs. 67%). This difference stood at 4% in the comparison areas (79% vs. 75%). This supported their hypothesis. Overall at parities 70% indicating a strong effect of son preference on fertility in matlab. When women had at least 2 sons at parity 3 and 3 sons at parity 4 they generally did not have any additional children. If they had no daughters at these parities however they tended to have additional children. this indicated a preference for at least 1 living daughter.