Proximate determinants of fertility

Abstract

The Bongaarts and Potter model of fertility determinants was adapted to the African context: the effects of marriage patterns contraceptive use postpartum infecundability and primary sterility. The effect on total fertility represented by births outside a union was included in the model such that the model captured the fertility inhibiting effect of union patterns and the fertility promoting effect of sexual relations outside of unions. In Mali Ondo state in Nigeria and northern Sudan most fertility occurred within unions. However in Botswana Liberia and Uganda substantial fertility occurred outside of unions. Abortion was not included due to gaps in data availability. In Africa postpartum abstinence and prolonged breast feeding have been strong influences on fertility decline and birth spacing. Contraceptive prevalence has been low in Africa with the exception of Botswana Kenya and Zimbabwe. Sterility has not been a reliable measure. A comparison of Kenya Senegal Sudan and Ghana had shown that postpartum infecundability was the greatest fertility inhibiting factor for all four countries. In Kenya contraceptive use also has had an impact on fertility change. Changes in marriage patterns were more unique to Senegal and Sudan as determinants of fertility decline. Ghana did not show evidence of changes in determinants or fertility decline. Data analysis was based on the Demographic and Health Surveys conducted during the 1980s in 12 sub-Saharan African countries and World Fertility Survey data during the 1970s for Ghana Kenya Senegal and northern Sudan with its sizeable Arab/Muslim population. The model showed total fertility rate being equal to the product of total fertility an index of marriage an index of contraception an index of postpartum infecundability and an index of sterility. The results for the mid-1980s reflected varying fertility from 4.9 in northern Sudan to 7.4 in Uganda. Bongaarts and Potter models have indicated an average proximate determinant estimate of 15.3. Among the 16 African populations the average was 14.1 which would indicate a 1-birth-lower fertility. Caution was urged in accepting this study model as reliable.