AN ANALYSIS OF THE PROXIMATE DETERMINANTS OF FERTILITY IN SUB-SAHARAN AFRICA WITH A FOCUS ON INDUCED ABORTION

Background: The fertility transition in sub-Saharan Africa (SSA) has been slower than in other regions of the world largely due to lack of rapid increase in contraceptive prevalence and the associated decline of unmet need for family planning. To study how fertility has changed across SSA, this dissertation will analyze the contributions of the proximate determinants of fertility to overall fertility decline by country, and undertake focal studies on one of the main proximate determinants – induced abortion. Methods: The data used for the analysis in this dissertation comes from the Demographic and Health Surveys. A total of 105 surveys covering 39 countries are used from SSA and seven non-SSA comparator countries. The first analytic paper calculates each of the proximate determinants using Bongaarts’ framework (1978, 1982) and decomposes change in fertility by country by the four major determinants. The second paper uses a country fixed effects multivariate linear regression model to determine the association between unmet need for family planning and induced abortion for all countries of SSA with available data, testing two different estimators of total abortion rate. The third analytic paper uses regression discontinuity design to isolate the impact of change in the abortion law in Ethiopia in 2005 on the incidence of reported pregnancy termination. Results: Most of the decline in fertility in SSA Is attributable to increases in the proportions of women unmarried, and to a lesser extent, increases in contraceptive use. A positive association exists between unmet need and total abortion rate for countries of SSA though the pathways inherent in this relationship are still unclear. And finally, the regression analysis in the third paper shows that the incidence of reported pregnancy termination neither increases nor decreases as a direct result of the change in abortion law in Ethiopia. Conclusion: Family planning programs should be further strengthened to reduce unmet need. Data on induced abortion need to be better collected and/or estimated. Loosening of abortion laws do not necessarily lead to greater uptake of abortion