Contraception and unintended pregnancy: explaining the unexpected relationship in Sub-Saharan Africa using demographic and health survey data

Introduction A body of scholarship has demonstrated that contraception enables women and couples to avoid unintended pregnancy (delay next birth or have no birth). However, the relationship between the modern contraceptive rate and unintended pregnancy is not well understood, with a small body of evidence at the global level documenting counter-intuitive findings. Cross-sectional analyses of the relationship between modern contraceptive prevalence and the proportion of unintended pregnancies in developing countries show an unexpected positive correlation. Methods Using Demographic and Health Survey data from 271 country surveys, we examined this relationship when disaggregated by major world regions. We first conducted the conventional cross-section contraception-unintended pregnancy relationship by fitting a simple linear regression (ordinary least squares) and deriving correlation coefficients for world regions. We then go further to conduct a longitudinal analysis of pooled data using random and fixed effects models. Results We found that the counterintuitive relationship is only held for Sub-Saharan Africa (SSA). Using correlation and regression coefficients (slope) of countries by stage of the fertility transition in SSA, we found that the counter-intuitive relationship is very strong in countries that are lagging or at early stages of the fertility transition. The correlation and regression results for different groups of countries at different stages of fertility transition show a clear pattern towards expected association as countries progress in the transition in SSA. Conclusions We conclude that the unexpected relationship between the proportion of unintended pregnancy and modern contraception is only temporary in the early stages of the fertility transition when the demand for family planning is higher than the ability of the health system to satisfy it.