CONTEXTUAL DETERMINANTS OF FERTILITY TRANSITIONS AMONG BLACK SOUTH AFRICAN WOMEN: A MULTI-LEVEL ANALYSIS

Abstract

Fertility declines (transitions) and their contextual determinants are concerns to the government and other stakeholders in South Africa. These are so, especially among the Black South African women of childbearing age. The South African Demographic Health Survey (SADHS) 2016 data was used to examine the role of three hierarchical layers of variables (individual, household and community level characteristics) in determining fertility transitions among Black South African women of childbearing age. Based on the Social-Ecological (SEM) and the Easterlin’s micro-economic models, the chi-squared test and multilevel logistic regression were performed at the bi-variate and multivariate levels of analysis. The multilevel logistic regression was performed using the generalised linear and latent mixed model (GLLAMM) to obtain fixed and random effects. Findings suggest that close to half (48.1%) of these women had low fertility levels (1-2 children) in South Africa. Those in rural areas had a higher fertility level (4 plus children), compared to those in urban areas. Factors such as mother’s age, employment and wealth status, owning a house with water and electricity, access and distance to health facilities and workers, etc., were strongly associated (significant) at different hierarchical model levels (p<0.05). Results of random effect revealed a non-existance (0.00%) of variations in their log odds of predicting fertility transitions across the communities (clusters/layers). The study recommends that these findings be considered in all programme and policy developments around the issue in South Africa.