Socio-economic status fertility preferences and contraceptive change in sub-Saharan Africa.

Abstract

Fertility has declined in several sub-Saharan African countries since the 1970s primarily (but not exclusively) as a result of the increased practice of contraception. Although this is highly probable there is considerable uncertainty about the forces underlying the increase in contraception. Among the unresolved issues are the causal contributions of changes in fertility desires and of changes in socio-economic factors such as schooling and rural-urban residence. In this paper we analyze the sources of increase in contraceptive prevalence in seven sub-Saharan African countries in the period from the 1970s to the 1990s using World Fertility Survey and Demographic and Health Survey data. Using a regression decomposition and change in prevalence is attributed to change in explanatory variables (composition component) and change in rates of contraceptive use within subgroups defined by the explanatory variables (rates component). The explanatory variables consist of two fertility preference indicators and measures of schooling (female and male) type of place of residence and occupation. The rates component dominates in six countries - Cote dIvoire Kenya Senegal Tanzania Uganda and Zambia - whereas in Ghana the composition component dominates. Some of the effects of the socioeconomic variables operate through changes infertility preferences and after taking these indirect effects into account the independent contribution of changing preferences falls far short of the contribution of the socio-economic variables in all seven countries. The results provide some support to the argument that a precondition for fertility decline in Africa is a transformation of fertility demand and that changes in socio-economic factors (such as schooling and urban-rural residence) are likely to be driving forces. But the analysis also reveals that significant increases in contraceptive prevalence can occur through satisfying existing demand for fertility control in the absence of changes in either fertility desires or major socio-economic factors. These findings lend support to the broad-based social development policy as well as more focused reproductive health programmes. (authors)