A flexible model for reconstructing education-specific fertility rates: The case of sub-Saharan Africa

<b>Background: </b>Accurate and harmonized estimates of education-specific fertility rates are crucial for understanding the past and projecting the future human population. Yet fertility estimates derived from demographic surveys that collect detailed fertility histories often do not align with the reliable and widely used United Nations (UN) World Population Prospects. This inconsistency means that the choice of data source can affect research outcomes on fertility trends.<br/><br/><b>Objective:</b> We combine the patchy Demographic and Health Surveys (DHS) data and the UN total fertility rate (TFR) estimates to create three harmonized datasets of education-specific TFRs for 36 sub-Saharan African countries, with different degrees of consistency with the UN TFR.<br/><br/><b>Methods: </b>We develop a flexible Bayesian hierarchical model that reconstructs education-specific fertility rates by combining the DHS data and the UN TFR estimates.<br/><br/><b>Results: </b>We provide time series of education-specific TFR quinquennial estimates between 1980 and 2014 for 36 sub-Saharan African countries. We present three model specifications that provide the users with fertility estimates that differ in their degree of consistency with the UN TFR.<br/><br/><b>Conclusions: </b>The model estimates show significant variation across countries, leading to divergent fertility trends from 1980 to 2014, mainly in levels but sometimes also in direction.<br/><br/><b>Contribution:</b> Our flexible modelling framework offers model specifications that suit different conditions and can obtain results consistent with stakeholder needs: consistent with but not identical to the UN, fully consistent (nearly identical) with the UN, and consistent with the DHS. Further, our estimates of education-specific TFR can be used to analyse and forecast fertility trends and their contribution to population change in 36 sub-Saharan African countries.